ARIZONA COTTON
PESTICIDE USE DATA: OPPORTUNITIES AND
PITFALLS
G. K. Agnew
and P. B. Baker
Pesticide
Information and Training Office
and G. B.
Frisvold
Agricultural
and Resource Economics
University of
Arizona
Tucson, AZ
Abstract
Pesticide use reporting systems are becoming more
widespread. The data gathered by these
systems has enormous potential for a wide range of applied research. Arizona's pesticide use reporting system
illustrates both the opportunities and pitfalls that will arise with the
increased use of reporting systems. The
data available provides a detailed and extensive picture of pesticide use in
major Arizona crops like cotton. With improvements
it could be the cornerstone of research that benefits producers and the
environment. As it stands, however, the
Arizona pesticide reporting system lacks fundamental information essential to
make it a useful data source. Summaries
illustrate both the strengths and weaknesses of the Arizona reporting
system. Preliminary research results
provide an example of the potential power of the dataset.
Introduction
Pesticide Use reporting is not going to go
away. California has required full
agricultural pesticide use reporting since 1990 (http://
www.cdpr.ca.gov/docs/pur/purmain.htm). Arizona has required full reporting by
custom agricultural applicators since the 1960s. New York state passed a bill in 1996 and started a program in
1997 to quantify pesticide use (http://pmep.cce.cornell.edu/
regulation/psur). Presently Oregon is
developing a reporting system and Wisconsin recently passed a bill mandating
development of a pesticide reporting system in the near future. Moreover, state mandated reporting systems
are only part of the story. In Arizona,
an estimated 50% of vegetable producers participate voluntarily in a private
pesticide use reporting program to provide processors with exact information on
potential residues on produce (Janzen 1999).
All of these reporting systems exist as a result of combined pressures
from consumer and environmental groups, state regulatory bodies and food
processors. All of these use reporting
systems are separate from the National Agricultural Statistics (NAS) surveys
which provide overall pesticide use statistics based on statistical samples
primarily for regulatory purposes.
There is no reason to believe these pressures will diminish in the near
term.
Arizona's pesticide use reporting system is
unique. It combines the detailed usage
information of the California system with accessibility and ease of use. While the system covers more than just
commercial applications for the whole state, it is easily manipulated with
desktop database software. This has
facilitated the development of useful applications for this kind of extensive
data and consequently stimulated discussion of just how this data can be
optimally used and collected.
Cotton is Arizona's most widely planted and most
valuable field crop. Combined with the
relatively intensive nature of cotton pest management, cotton pesticide use is
disproportionately represented in the Arizona pesticide use database. Furthermore, much of the present and planned
research that takes advantage of this data is focussed on cotton. This paper has multiple purposes: Discussing pesticide use reporting in
general; Summarizing recent trends in Arizona cotton pesticide use; Illustrating the power of Arizona's
pesticide use database; and using our
experience of working with this database as a vantage point from which to
consider the options available to future pesticide use reporting systems.
Methods
Use
Reporting: The Basics
General summaries of pesticide use are confounded by
the diversity of pesticide products. Products are usually categorized by active
ingredient (AI) for general summaries.
This allows for some aggregation of the numerous pesticide products used
in agriculture. Liquid and dry
formulations of differing strength can all be grouped together. Within a single crop and AI there is
relatively little variation in application rates compared to the variation
across crops within an AI or across AIs.
With a pesticide product database
with conversion factors it is quite simple to summarize pesticide use by
AI.
By pounds
of AI
While it is natural to want a single measure that
describes all agricultural pesticide use, summing pounds of pesticide use
across AIs is not a useful proposal.
Recommended label rates vary from .01 pounds per acre for some
pyrethroid products (ie. Karate) to over 100 pounds per acre for some
nematicides (Telone II). Combining
usage statistics from applications of these two products would effectively lose
the information on the pyrethroid while overstating the importance of the
heavier nematicide. This practice
becomes particularly misleading when observing trends over time. Marked increases or decreases in an overall
usage measure could be the result of swings in a single product. Conversely, major increases or decreases in
the use of the pyrethroids could easily be masked.
Treated
Acres
Measures of treated acres is another approach to
summarizing pesticide use. NAS has
traditionally reported the percentage of fields treated and the average number
of treatments per field. This kind of
information has become particularly important in the last three years with the
passing of the Food Quality Protection Act (FQPA). In the FQPA regulatory process, in the absence of actual usage
data, regulators use 100% of acres at maximum label applications.
Reporting acres treated is a relatively simple process
for the producer and is information that is easily stored in a database. It is more difficult to design and implement
a system that determines which acres
are being treated so as to determine multiple treatments. Ideally each field
would have an ID number that would allow for identification of the field
through the season. California has attempted to include this in its reporting
system. In Arizona, no attempt has been
made to track fields. With multiple
crops and changing field size and location year to year field tracking
increases the complexity of use reporting substantially.
This seemingly small detail is actually a major
limitation in the Arizona reporting system and a potential weak link in any
system. Without field identification it
is impossible to know which acres have been treated how many times.
Application
Intensity
An alternative measure is an intensity measure which
determines how many applications would have been made if every acre had been
treated equally. This is simply the ratio
of treated acres to planted acres.
While this statistic gives a rough idea of usage, it masks any variation
within the area considered. For
instance, at the state level in 1998, Arizona had an application intensity for
insecticides of approximately five implying all fields were sprayed, on
average, five times. At the county level, however, intensities range from .22
to 6.22. Clearly usage patterns vary
across the state. If the area
considered was pared down to the field, which would necessitate field tracking,
then the result would be the number of treatments on each field
In the Arizona 1080 reports, fields are identified
by range, township and section, which generally limits the potential cotton
acreage to 640 acres. Within the
section there is no way, however, to know how many acres are actually planted
in cotton. Reports of five ten acre
applications could be one field receiving five applications or five fields
receiving one application or any combination in between. For research discussed later in this paper,
we have been able to obtain section level cotton acreage from an Arizona
growers group which allows for section level intensity measures. While variation is unquestionably being lost
within the section level data, without field tracking this is the only way that
section level data can be made useful for statistical analysis.
Only a true intensity measure allows for comparisons
of usage across large areas and across time.
Ultimately, different areas must be compared in terms of percent of
acres treated. Under the present system
this can only be done at the county level or where other measures of cotton
acreage are available. Over time,
percentages are necessary to control for changes in planted acres. Any overall gross measure of Arizona
pesticide use in cotton over the last five years would show usage falling
precipitously. This statistic would
reflect the steep decline in acres planted to cotton in the state. The "intensity" measure that
normalizes treated acres treated with planted acres provides a reasonable way
of comparing across years.
Application
Rates
Despite the lack of field tracking, the Arizona use
data provides a means of knowing actual application rates for pesticide
products in the field. Pesticide labels
only provide a range of recommended rates for a certain product. The same product might have a quite wide
range of recommended rates for different target pests. As mentioned before, in the absence of
better data the default assumption for regulatory purposes has been full label
rates.
Actual pesticide use rates only necessitate a
measure of acres treated by a particular tank of pesticides. Arizona's reporting system requires
reporting of both pounds of product to be applied and acres on which that
pesticide will be applied. A simple
conversion from pounds of product to pounds of AI allows for comparison of
application rates within an AI.
Mixed
Applications
Finally, usage rates are further complicated by the
possibility of mixed applications of two or more AIs. Pesticides are widely used in combination and
yet little information about this practice is available. While in general the practice of tankmixing
products may be more common with herbicides, it has also been a common practice
with insecticides in Arizona in recent years.
In 1995, the worst year of a whitefly infestation, 488 combinations of
up to five different AIs were recorded.
These multiple AI tankmixes complicate both the application rate and
applied acres measures.
Data
Pesticide use reporting is not complete in
Arizona. Much like California's system
prior to instituting full reporting, Arizona requires the reporting of only
certain kinds of applications. In addition
to the omission of field-tracking, this is clearly a weakness of the database. Recognizing these weaknesses, it is still
possible to make good use of the data that is available.
Arizona mandates reporting of pesticide applications
by commercial applicators, the applications of pesticides registered under
section 18 registrations and certain applications of pesticides included on the
Arizona Department of Environmental Quality (DEQ) Groundwater Protection List
(GWPL). Anecdotal evidence indicates
that there is also voluntary reporting of unregulated pesticide applications
Commercial applicators have a strong incentive to
comply with reporting regulations. They
can lose their state license if they do not follow proper procedures. In Arizona, commercial applicators play a
major role in pesticide application because of the importance of aerial
applications of pesticide, all of which are done by commercial operations.
Section 18 registrations have been important for
tracking new chemistries as they enter into cotton production. The insect growth regulators (IGRs) (Knack
and Applaud) are the most recent examples of section 18 registrations that
should have full reporting in the database.
Once again, the incentive for producers to report is relatively high
because continued registration of the product is dependent on full reporting. On the other hand, the potential penalty for
an individual producer is small relative to the commercial applicator and
unfamiliarity with he reporting system may lessen compliance.
Finally, the reporting of GWPL applications, which
should be complete for those AIs on the list, is actually difficult to
quantify. The lack of a visible
regulatory presence for the GWPL and relative lack of producer incentives make
this aspect of the Arizona use reporting system potentially unreliable. Because the GWPL applies only to
soil-applied applications this uncertainty affects reports of herbicides and
nematicides more than insecticides and defoliants.
Results
Using the
Arizona Pesticide Database
The limitations of the Arizona reporting system
determine where the data can be most useful.
At a minimum, reported applications provide a lower bound for actual
applications in the state. They also
provide hard evidence of the range of practices being used by producers across
the state. Furthermore, the reporting
system has been in place without serious structural changes since 1993 when the
GWPL was included. Thus trends over
time should reflect actual trends within the group reporting applications.
In the scenarios where reporting can reasonably be
assumed to be full, or almost full, more involved hypotheses can be made. The reporting of applications of IGRs should
be relatively complete and thus summaries of their use should represent their
actual use in the state. In general,
the use of aerial applications of insecticides in cotton leads us to believe
that a high percentage of insecticide applications in mid and late season
cotton are included in the dataset.
In 1998, there were 2.1 million application-acres on
265,900 acres of Arizona cotton.
Figure
1 shows that application-acres have declined to less than half of the level of
the early 1990s. Maricopa and Pinal
counties represent the majority of the treatment acres. Figure 2 shows application intensity --
application-acres normalized by the acres planted in cotton for both the state
and the individual counties. At the
state level, even controlling for the steady decrease in acres planted to
cotton, the number of pesticide applications has declined. This is also true for the major production
counties of Pinal and Maricopa.
1995 represents the high point in application
intensity for both the state, at 14.9 applications per acre, and the major
producing counties. This was the year
before IGRs were available and whitefly infestation was high. Where infestations occured, a high number of
insecticide applications were made to avoid potential yield loss and reduced
lint quality. This was also a period of
heavy use of tankmixes which, because application-acres are included for each
separate active ingredient, will inflate the intensity measure.
Figure 3 and figure
4 show the reported application-acres
and application intensity broken down into type of pesticide. It is easy to see that insecticides dominate
the database. This result is not
unexpected as the number of applications of insecticides will commonly dwarf
the number of applications of any other single category.
It is also important, however, to recognize how the
limitations of the database might manifest themselves in these numbers. Herbicides are clearly severely undercounted
in the pesticide use reports. Producers
frequently apply their own herbicides at or before planting and at layby. Thus the only non-voluntary incentive to
report would come from the GWPL. The
sharp rise of reports in 1993 with a decline thereafter is consistent with the
publicized onset of the new regulatory program and the subsequent decrease in
awareness thereafter. In fact anecdotal
evidence indicates that herbicide use in general is on the increase.
Defoliant usage, which is probably well represented,
as it is frequently aerially applied, appears to be steady. Plant growth regulator usage (Pix), at 65%
of the acreage in 1998, is 50% higher than any previous year.
Figure 5 shows insecticide intensity by Arizona
counties. As expected Maricopa and
Pinal counties are near the top in terms of application intensity. During 1995
these counties experienced widespread whitefly infestation. Interesting in this graph is the disparity
between La Paz, Yuma and Mohave counties. All in western Arizona along the Colorado river, these three
counties appear to have very different usage patterns through the 1990s. Application intensities in Mohave county for
all kinds of pesticides are consistently lower than other counties. This might indicate a generally lower
reporting rate rather than a lower
level of usage.
Figure 6 shows herbicide intensity in Arizona
counties. As mentioned before,
herbicide reporting is likely to be low.
Other than a few aberrations this graph still illustrates a general
decline in herbicide use reporting after highs in 93 and 94.
Figure 7 shows defoliant intensity in Arizona
counties. Relatively low intensities in
the eastern counties of Cochise and Graham are not a surprise both because
cooler fall weather assists the defoliation process and reporting is likely to
be lower from these areas. Once again,
Mohave county's relative low reported intensity appears to be an anomaly.
Figure 8 shows the plant growth regulator intensity
in Arizona counties. There is
considerable variation at the county level from year to year, perhaps as a
result of the weather dependant nature of plant growth regulator use. The overall intensity measure indicates the
general increase in plant growth regulator usage.
Figure 9 and figure
10 track the usage of the top ten
insecticides used in Arizona between 1991 and 1998. The general reduction in insecticide use is reflected as is the
whitefly infestation in the middle part of the decade. Fepropathrin (Danitol) is a dramatic example
of an AI, heavily promoted for whitefly tankmixes, that has seen reduced usage
with the registration of IGRs.
Tables 1, 2 and 3 provide 1998 use statistics by
pesticide type, with preliminary 1999 use statistics collected through
September 1st. Preliminary
1999 numbers indicate that pest pressure was low. The increased use of Roundup Ready cotton is indicated in a 100%
increase in glyphosate usage. Continued
use of Bt. cotton explains the continued low usage of gossyplure, a pink
bollworm pheremone that has been heavily used in Arizona.
An
Application: Adoption of IGRs
A research project was developed to explore the
potential power of the Arizona pesticide use data. The limitations of the
Arizona use reporting system were taken into consideration. A study of the adoption of IGRs and the
subsequent effect on pesticide applications takes full advantage of the
strengths of the use database while sidestepping the acknowledged limitations.
The IGRs pyriproxyfen (Knack) and buprofezin
(Applaud) were granted section 18 status beginning in the 1996 growing season
for the purpose of combating whiteflies.
Producers were limited to one application of each product and reporting
was mandatory. Thus, it is reasonable
to assume that IGR reporting in 1996 was complete within the limits of
regulatory compliance.
In 1995, prior to the registration of IGRs, in
problem areas producers treated as many as 12 times to minimize whitefly damage
(Dennehy and Williams, III, 1997).The database should include the majority of
the whitefly applications because whitefly pressure primarily occurs after the
cotton canopy has closed over the rows, necessitating commercial aerial
application of whitefly-targeted insecticides.
Discussions with producers and extension agents indicate that
specialized equipment needed to treat later season cotton from the ground is
the exception and that in many areas, heavy irrigation schedules would make use
of this equipment impossible.
It is important to identify applications
specifically targeting whiteflies. This
can be accomplished by focussing on certain tankmix combinations. As a result of grower experience with, and
extension research on, the whitefly infestation in Arizona, by 1995 the
efficacy of pyrethroid-organophosphate combinations was already widely
recognized in 1995 (Dennehy et al. 1995).
Explicit insect resistance management (IRM) guidelines were developed
recommending that non-pyrethroids be employed against other pests to maintain
efficacy of pyrethroids singly and pyrethroids synergized by an organophosphate
or endosulfan (Ellsworth and Diehl, 1995).
This study utilizes acreage data on the IGRs, a
variety of tankmix combinations that include combinations of active ingredients
indicated in extension publications (Ellsworth et al, 1994, Ellsworth et al,
1996) and an overall tankmix aggregate.
The most commonly used whitefly tankmix in 1995 is an
acephate-fenpropathrin (Orthene-Danitol) combination The aggregate tankmix acreage was considered because so many
different permutions of potential whitefly active ingredients were used in
1995. There were 488 different tankmix
combinations including up to five active ingredients. 280 of these combinations
were included in the aggregate tankmix variable as likely whitefly
applications. Table 4 shows the most
common whitefly tankmix combinations.
In the tankmix variable, all combinations include at least one
pyrethroid and a non-pyrethroid. We removed combinations including the pink
bollworm pheromone gossyplure, whitefly-specific imidicloprid and all
non-cross-family mixes (ie. two organophosphates, chlorpyrifos and acephate
(lorsban and orthene) not a combination deemed effective against whiteflies).
For this analysis, the Arizona Cotton Research and
Protection Council (ACRPC) provided data on cotton acres at the section
level. As discussed earlier, an acreage
measurement is necessary to normalize application-acres into a meaningful
measure of application intensity. This measure of mean applications per section
masks variation within a section but makes it possible to use the unusually
disaggregate section level data. A
section is 640 acres, while a third of Arizona cotton farms are 500 acres or
more (USDA, 1999). These farms
accounted for three-quarters of Arizona's cotton acreage in 1997. In this way, each section of the state where
cotton is grown becomes an observational unit.
It makes it possible to construct a large, geocoded database on
pesticide use intensity with between 1634 and 2157 observations per year
between 1995 and 1998. This study makes
use of a section-level database to examine (a) factors explaining IGR adoption
and (b) how adopters of IGRs altered their overall insecticide use to control
whiteflies.
Preliminary findings indicate that IGR adoption can
be explained to a large extent by location effects. Adoption was also more likely on sections where an index
measuring whitefly susceptibility to synergized pyrethroids was low and where
whitefly applications were larger the previous year. Adoption was inversely related to local population density. On sections where growers adopted IGRs,
expenditures on synergized pyrethroid and other tank mix applications fell by
$62.52 per acre. On sections with no IGR adoption, tank mix expenditures fell
less, by $44.37 per acre. On adopting
sections, net costs of controlling whiteflies fell by $29.62 per acre, or by
over $11,000 per farm. (See Adoption of
Insect Growth Regulators in Arizona Cotton:
Determinants and Economic Implications, in the 2000 Beltwide
proceedings)
Discussion
With the historically limited available information
on pesticide usage the focus of research has always been on relatively simple
characterizations of use patterns. With
a use reporting system like the Arizona L1080 form the possibilities expand
substantially. The forms include the
date of application, whether the application was made by ground or air and
starting this year, the target pest.
All of this data coupled with the power of spatial analysis/GIS mapping
and statistics offers endless opportunities for research directly useful to the
cotton producer.
Work is under way by entomologists to utilize this
data to better understand the ecology of pest populations.. Pest Control
advisors will soon use this information to improve IPM decision-making and
resistance management. Plant pathologists have already taken advantage of GIS
mapping of pesticide use patterns to better understand nematodes in
cotton. This kind of research is still
in its infancy and hold great promise for production agriculture right down to
the field level with precision agriculture.
Another obvious application for this data is further
support of the regulatory system so as to assist producers. When special, localized problems arise which
call for a section 18 or SLN
registration, the information necessary to determine the extent and severity of
the problem will be available.
Resistance issues can be substantiated with actual use data, along with
insect counts and susceptibility measures.
In another regulatory arena, pesticide use data is providing fact-based
alternatives to exaggerated default assumptions being used in reregistration
decisions and risk assessment as a result of the Food Quality Protection Act.
Summary
With the increased interest in pesticide use
reporting systems there is an opportunity for researchers to gain valuable data
for applied agricultural research. As
an established use reporting system, the Arizona pesticide use data represents
the opportunities and pitfalls of these systems. In its present form, the Arizona pesticide use reporting systems
provides useful data for detailed summaries of Arizona pesticide use and
limited statistical research. Results
show a general decrease in the use of insecticides in Arizona cotton and a
substantial economic impact of IGRs.
With the addition of complete coverage, field tracking and secondary
farm and producer data, the pesticide use data would be an even better dataset
providing a wealth of data for wide range of research agendas.
Funding
National Agricultural Pesticide Impact Assessment Program
and EPA Pollution Prevention Incentives for States Grant through the Arizona
Department of Environmental Quality
References
Dennehy, T. J., A. Simmons, J. Russell, and D.
Akey. (1995). Establishment of a whitefly Resistance Documentation and
Management Program in Arizona.
Cotton: A College of Agriculture
Report, Series P-99.
Dennehy, T.J., and L. Williams, III. 1997.
Management of Resistance in Bemisia
in Arizona Cotton. Pesticide Science
51:398-406.
Ellsworth, P., and J. Diehl. (1996, revised
1997) Whiteflies In Arizona: Insect Growth Regulators 1996. Arizona College of Agriculture extension
publication.
Ellsworth, P., and T. F. Watson. (1996). Whiteflies in Arizona: Pocket Guide '96. University of Arizona, College of Agriculture Cooperative Extension.
Ellsworth, P., L. Moore, T.F. Watson, and T Dennehy.
(1994) 1994 Insect Pest Management for
Cotton. University of Arizona, College of Agriculture Cooperative Extension.
Janzen, R. CDMS (Crop Data Management Systems). Personal communication, 1999.
