|
|
|
|
California
pesticide use reporting (PUR) data
The state of
California has a unique pesticide use reporting (PUR) system that
other states in the U.S. are only beginning to emulate. Most commercial
pesticide applications in the state are reported to the California
Department of Pesticide Regulation (DPR), a total of 2-3 million
records annually, with 192 million pounds of pesticide active ingredients
applied in 2000. No other state and no other region of the world
has such a comprehensive system for tracking pesticide use (although
Oregon will implement a similar system in 2002). The scale and comprehensiveness
of the PUR dataset make it a useful resource for researchers worldwide.
 About
the data
What
information is contained in the PUR data set?
What
types of pesticide applications are reported in California?
What
types of pesticide applications are NOT reported in California?
What
else is missing from the pesticide use reports?
How
is California pesticide use information used?
Errors
in the data
What
information does the PAN CA PUR web site report?
How
does PAN process the PUR data for use on this site?
Defining
acreage: Treated vs. planted acres
Methodology
for calculating planted acreage
Building
a unique location identifier
Using
the location ID to calculate crop acreage
Problem
crops
Evaluation
of the Accuracy of Acres Planted
About
the data
California
EPA's Department of Pesticide Regulation (DPR) is responsible for
collecting and maintaining the California PUR data set. Summary
reports on the data are available for download from DPR's web site
(1) . The raw data files (~500 Mb per year of data) are available
on CD-ROM in fixed-width or dbf formats and can be ordered from
DPR.
The state of
California has been collecting limited pesticide use data since
the 1950's, with full use reporting beginning in 1990. Additional
PUR history can be found on DPR's website (2) and in a document
produced by PAN Germany (3). The PAN Pesticide database includes
the California data from 1991 through 1999. We do not report data
for 1990 because the data for this first year are so poor that they
are generally not useful for analysis. The 2000 data are preliminary;
final 2000 data are expected sometime in 2002.
References:
- Pesticide
Use Reporting (PUR),
California
Department of Pesticide Regulation. Viewed on October 31, 2002.
- Pesticide
Use Reporting: An Overview of California's Unique Full Reporting
System, California
Department of Pesticide Regulation. Viewed on October 31, 2002.
- L. Neumeister,
Pesticide Use Reporting: Legal Framework, Data Processing and
Utilisation, Part One: Full Reporting Systems in California and
Oregon, Pesticide Action Network Germany (Hamburg, 2002),
in English. Download.
Viewed on October 31, 2002.
Top of page
What
information is contained in the PUR data set?
The
data collected by the California Department of Pesticide Regulation
include the identity and amount of the pesticide applied, crop or
other site treated, geographic location of the application (to the
square mile for agricultural uses; to the county level for nonagricultural
uses), method and date of application, acres planted and acres treated
for field applications, and operator identification codes.
Top of page
What
types of pesticide applications are reported in California?
The
California Pesticide Use Reporting system requires commercial applicators
of pesticides to report all pesticide applications to the California
EPA's Department of Pesticide Regulation (DPR). The reports are filed
by farmers (large commercial growers as well as small farmers who
market their goods), commercial agricultural pesticide applicators,
structural pest control companies, and commercial landscaping firms.
Top of page
What
types of pesticide applications are NOT reported in California?
Consumer
use of home and garden pesticides is not reported, nor are most industrial
or institutional uses, unless applications are performed by a licensed
pest control applicator. Using pesticide sales data, it is possible
to estimate unreported pesticide use at approximately 30% of total
pesticide use in the state.
Top
of page
What
else is missing from the California pesticide use reports?
Only the "active ingredients" in a pesticide are reported.
"Inert" ingredients, which can account for up to 99% of
the total weight of applied pesticide, are not included in the data
reported. Unfortunately, "inert" ingredients are not necessarily
inert. Many are known to have adverse health effects and may indeed
be pesticides in their own right. Eight inert ingredients are classified
as "Of Toxicological Concern" and another 75 are "Potentially
Toxic." Examples of toxic "inerts" include formaldehyde
(a pesticide and a carcinogen), chloropicrin (a pesticide and a
potent neurotoxin), o-cresol (causes genetic damage), ethoxylated
p-nonylphenol (an endocrine disruptor),
o-phenylphenol (a carcinogen), toluene (a developmental toxicant),
and xylene (a neurotoxin). While
the identity of specific "inerts" in a particular pesticide
product is not available to the public, it is possible to estimate
the additional percentage of inert ingredients released into the
environment using California Pesticide Use Reporting (PUR) data.
In 1998, a total of 215 million pounds of pesticide active ingredients
and 134 million pounds of inert ingredients (an additional 62% more
chemicals) were applied in California. Pesticide products used in
urban areas typically contain more inerts than agricultural-use
pesticides. To find out more about inerts, see Toxic
Secrets: "Inert" Ingredients in Pesticides 1987-1997.
Other information
that is desirable but not reported includes: the target pest, whether
a crop is being grown organically or is genetically modified, planting
and harvest dates for crops, and identification of special, "sensitive"
sites such as hospitals and schools.
Top
of page
How
is California Pesticide Use Information Used?
Pesticide use
information is the necessary foundation for all aspects of pesticide
regulation, risk reduction, environmental protection and pest control
research. Accurate and detailed pesticide use information such as
that gathered through a use reporting system benefits a diverse
clientele, from pesticide applicators and pesticide manufacturers
to researchers, regulators, growers, farmworkers and community members.
In particular,
the data can be used to assess and improve pest management strategies,
provide crop acreage counts and market forecasts for farmers, and
track trends in pesticide use. Researchers and regulators use the
data to investigate the effects of pesticide use on air quality,
groundwater and surface water quality, human health, and endangered
species. A number of specific examples are detailed in reference
1.
Pesticide use
information also serves the community right-to-know by providing
community members with information on what types and amounts of
pesticides are used near their homes so they can take appropriate
action to prevent further exposures. One of the drawbacks of the
California PUR system is that this information is not available
in a timely way unless it is obtained from the County
Agriculture Commissioner. However, even knowledge of typical
pesticide applications by a particular grower has led to negotiations
between farmers and community members in an attempt to reduce pesticide
use near homes, schools, and parks. Some of these negotiations have
resulted in "good neighbor" agreements where farmers voluntarily
agree to reduce the amounts of toxic pesticides applied near such
sensitvie sites or to farm those areas organically and to notify
neighbors when pesticide applications are scheduled to occur.
References:
- L. Neumeister,
Pesticide Use Reporting: Legal Framework, Data Processing and
Utilisation, Part One: Full Reporting Systems in California and
Oregon, Pesticide Action Network Germany (Hamburg, 2002),
in English. Download.
Top
of page
Errors
in the data
Errors
contained in the California Pesticide Use Reporting (PUR) database
include inexplicably high or low application rates, inaccurate acreage
counts, incorrect site ID numbers, errors in the percent of active
ingredients in a particular pesticide product and others. Some of
these errors have little effect on pesticide use statistics while
others are more problematic.
In the late
1990s, DPR developed error-checking routines that flag probable
errors when application rates are too high. These records are flagged
as 'outliers' and and these data points should be removed when working
with the data. While the number of records in error flagged by this
method is approximately 0.4% of the total number of records, errors
in total pounds of pesticides used are quite large---approximately
2–5% of the total pounds of reported use in the state---on
the order of 4–10 million pounds each year. DPR's error-checking
routines are improving each year, but to adequately address these
problems, better point-of-entry error checking is needed. DPR is
in the process of developing a web based data entry system which
should alleviate many of these problems. See DPR's methodology for
flagging outliers (1) for more details and their review of data
quality (2).
Two general
types of errors exist in the PUR data. The first is related to chemical
use, the second related to acreage reporting. To date, most of the
error correction analysis has focused on correcting chemical application
rates, specifically by flagging chemical application rates that
are above normal. For this reason, the reported pounds of active
ingredient is generally accepted as being an accurate number , within
the constraints of a system that relies on hand data entry.
The most significant
errors remaining in the PUR data are in the reported number of "acres
planted" for some crops, and work on correcting these errors
is now beginning to be seen as a priority. Because
use reporting in California is legally required, the PUR data is
potentially a highly accurate source of crop acreage statistics.
Currently this data is compiled for the entire U.S. by the US Department
of Agriculture (USDA) National Agricultural Statistics Service (NASS)
at a cost of over US $100 million per year. Using the PUR data to
generate crop acreage estimates has the potential for greater accuracy
and can be done at a fraction of the cost of current methods. DPR
is presently piloting a project in several counties to utilize satellite-based
Geographic Information System (GIS) to specify a field location
by latitude and longitude, which should improve future data quality
in this area.
In the meantime,
because acreage data in the PUR dataset still contains errors (some
of them quite large), users should evaluate the raw PUR data before
relying too heavily on summary acreage data generated from the PUR.
See the methodology section below for
a more complete discussion of PAN's data processing techniques and
error correction routines.
Finally, reported
pesticide use is only one way to track pesticide use in California.
DPR also collects pesticide sales data, which can be ordered in
printed form from DPR. The total amount of pesticides sold is approximately
50% higher than total reported pesticide use. Since 1995, the discrepancy
between sales data and reported use data has grown. There are several
possible causes of this discrepancy: 1) Increased consumer and institutional
use, 2) lack of full reporting by pesticide users, 3) errors in
the sales database, and/or 4) chemical stockpiling.
References:
- Pesticide
Use Reporting: An Overview of California's Unique Full Reporting
System, Appendix A, California
Department of Pesticide Regulation. Viewed on October 31, 2002.
- L. Wilhoit,
M. Zhang, and L. Ross, Data Quality of California's Pesticide
Use Report, Final Report to the California Department of Food
and Agriculture for Contract Agreement No. 0241, California Department
of Pesticide Regulation. Download.
Viewed on October 31, 2002.
Top of page
What
information does the PAN CA PUR web site report?
The
Pesticide Action Network has summarized the California PUR data in
an easily searchable form by crop, chemical, and geographical area.
The resulting data provides information on pesticide use for all combinations
of pesticide, crop, and geographic area. The following information
is reported:
- CA
DPR Chem Code: The California Department of Pesticide Regulation
(DPR) assigns a unique chemical code number to serve as an identifier
for a particular pesticide active ingredient or mixture of active
ingredients. For more, see about
CA DPR chem codes. For information on use of a particular
chemical between 1991 and 2000, select Search CA Pesticide
Use and use the pop-up menu to select a chemical of interest.
[Note: You must also select a geographic area and a crop or site.]
The chemical code number will automatically be entered in the
search box when you select a chemical from the list. The pop-up
menu shows a partial list of the available chemicals. The chemicals
contained in the list are a combination of the most commonly logged
requests on the PAN website and the highest-use chemicals in California.
If you wish
to look up information about chemicals not in the pop-up list,
you can select "Look up a chemical not on this list"
and a small window will open. Here you can search for the pesticide
to find the CA Chem Code, then return to the CA PUR search window
and type the chemical code into the pesticide search box.
- Crop
or site:
The California PUR data provides information on pesticide use
by crop or site. The California Department
of Pesticide Regulation (DPR) assigns a unique CA site code
number to serve as an identifier for a particular crop or
site. To find pesticide use information for a particular crop
or site for the most recent data year, select the crop or site
of interest from the list of crops.
For
information on pesticide use on a particular crop between 1991
and 2000, Search CA Pesticide Use and use the pop-up menu
to select a crop or site. The site code number will automatically
be entered in the search box when you select a crop from the list.
Since full pesticide use reporting began in 1990, pesticide applications
to a total of 262 different crops or sites have been reported.
At present, only crops with over 5,000 acres planted in California
in 1999 are included in the pop-up list, but the data for any
crop is accessible if you know the site code. For crops not in
the list, you can find the site code numbers by going to the lists
for total pesticide use in California from the top level CA
PUR Search page. U.S. EPA also assigns a U.S. site code
number to each crop as a unique identifier.
- California
Department of Food and Agriculture (CDFA) Commodity Category:
CDFA evaluates commodities by groups of similar crops. These groups
include: Vegetables and Melons, Fruits and Nuts, Field Crops,
Spices and Herbs, Nursery Products, Forestry, Livestock, and Other
Agriculture.
- Use
Category :
PAN has divided the crops/sites to which pesticides are applied
into Agricultural and Nonagricultural uses.
- Geographic
area: At present, you can select the county or counties in
which you wish to find out about pesticide use by using the pop-up
menu. Selecting a geographic area (region, county, or all California)
inserts a CA County Code number or numbers into the search box.
The California Department of Pesticide Regulation (DPR) assigns
a unique county code number to serve as an identifier for each
county Selecting "All California" provides results for
all counties in California.
In the future, we plan to include information on pesticide use
by square mile, using the meridian/township/range/section spatial
identifiers contained in the PUR data. This information will probably
be presented in the form of a Geographic Information System (GIS)
map.
- Gross
pounds of active ingredient: This number includes the total
pounds of the pesticide active ingredient applied to the selected
crop or site in the selected geographic area. Inert ingredients
are not included in this total. For some agricultural applications,
the gross pounds may be more than the pounds applied to acreage
because pesticides are used in processing the crop after harvest
or are used only on selected sites in the field (planting holes,
etc.).
- Application
rate This number includes the total pounds of the pesticide
active ingredient applied to the selected crop or site in the
selected geographic area. Inert ingredients are not included in
this total.
-
Application count: For agricultural pesticide applications,
this number refers to the number of applications in which the
selected pesticide was used on the selected crop or site in the
selected geographic area. This number is for applications to fields
only and does not include non-field applications. To be considered
a 'field application,' the 'unit treated' and 'unit planted' must
both be 'A' (for 'Acres'). For nonagricultural applications (structural
pest control, landscaping, etc.), the number of applications refers
to the number of individual units treated (homes, yards, apartments,
etc.).
- Number
of fields: For agricultural applications only. The number
of individual fields to which the particular pesticide was applied
in the selected geographic region. See methodology section below
for a more comprehensive discussion on fields.
Top of page
How
does PAN process the PUR data for use on this site?
The
raw PUR data from 1991 through 1999 contains over 30 million records.
To accelerate queries to the online database, we create slightly over
1 million summaries, one for every possible combination of chemical,
crop and county. Before
creating summaries, all records marked as outliers by DPR are excluded
from the raw data. PAN also does some data cleanup before summarizing
the raw data. The following sections describe the concepts and techniques
used in data cleanup.
Defining
Acreage: Treated Acreage vs. Planted Acreage
There are two
reported acreage figures in the PUR data: 'Acres Treated' and 'Acres
Planted'. The difference is best explained through an example:
If a farmer
has a 100-acre field and sprays 50 acres, then the Acres Planted
will be 100 and the Acres Treated will be 50. If a farmer sprays
50 of his or her 100 acres three times, then the Acres Planted will
remain at 100 acres, but now the value of Acres Treated will be
150. We label the total pounds of pesticide per Acres Treated the
'Treated Intensity'. When summarizing treated acres, all Acres Treated
figures are summed. If this is then divided by pounds of AI applied
to these acres, the resulting 'Treated Intensity' figure should
correspond with the application rate that is recommended on a pesticide
label.
PAN does not
currently apply any error corrections to the treated acreage numbers.
However, DPR's outlier routines look for unusually high application
rates (pounds of AI/ acres treated). This has the effect of excluding
records when either the pounds applied is very high or the acres
treated is very low. In addition, DPR applies a filter which rejects
Acres Treated values that are higher than those for Acres Planted.
The result is that Acres Treated is generally regarded as a fairly
reliable number.
The Acres Planted
figure is more problematic. We define Acres Planted as the total
acreage of a particular crop to which pesticides are applied in
a given year. This includes more than just harvested acreage (the
usual figure reported by the U.S. Dept. of Agriculture) because
any part of the crop that was treated is counted, not just what
was harvested. In the case of perennial crops such as trees and
vines, this means bearing plus nonbearing acreage. This number is
generally correct because most crops are treated with pesticides
at least once during the year. The Acres Planted number may include
fewer acres than are actually planted for crops that are
not treated with pesticides every year, such as forests, pastureland
and rangeland.
One problem
with the way pesticide use is recorded is that the value of Acres
Planted for crops planted in fall of one year and harvested in summer
of the next year (strawberries are a good example) is often incorrect.
As an example, consider the strawberry grower who plants 100 acres
in 97-98, 50 acres in 98-99 on a different plot of land,
and then 25 acres in 99-00 on the first plot of land. For the '98
reporting year, it will appear that the grower planted 150 acres
of strawberries. For the '99 reporting season, it will look like
the grower planted 75 acres. If the grower had instead used the
same plot of land all three years, it will just be confusing as
to how much of the land was actually planted in strawberries.
There are similar
problems with crops like lettuce or broccoli that have several plantings
in a single year. For accuracy, these crops should be counted once
for each planting and harvesting. Thus, a 20 acre field which is
planted with 2 lettuce crops per year should be counted as 40 planted
acres of lettuce. This does not occur at present.
There are
two important consequences of these problems for data processing.
First, to avoid over-counting acreage, in any given search, each
location to which pesticides are applied should only be counted
once unless the second instance of the location represents a new
crop. Secondly, because the plant and harvest dates are not reported
in the PUR data, using the PUR to calculate Acres Planted can be
difficult or impossible for these problem crops. In the PAN search
results, a comment is inserted for these kinds of crops to alert
the user to potential problems with these data.
See also Methodology
for Calculating Acres Planted, Problem
Crops, and Evaluation of the Accuracy
of Acres Planted for more information.
Top of page
Methodology
for Calculating
Acres Planted
Because the
State of California requires that all agricultural pesticide applications
be reported, and because pesticides are applied to nearly all cropland,
the PUR data set contains at least some record of nearly all cropland
in California. Each PUR record contains a grower ID, site locator
ID, site (commodity) code and the acres planted in that commodity.
With this data, total crop acreage (Planted Acres) can be determined.
This technique will give low estimates for sites that may not receive
any pesticide applications in a given year, such as forests and
pastureland.
The process of calculating
Acres Planted correctly begins with evaluation of specific identifying
information for each location to build a unique location identifier.
Building
a Unique Location Identifier
The California
Department of Pesticide Regulation (DPR) PUR data user documentation
indicates that locations can be uniquely identified by concatenating
the grower ID and site locator ID. However, due to variations
in how farmers report pesticide use, a more reliable location
identifier is a concatenation of the grower id, site locator ID,
and site code, a value we call the location ID. If each of the
component data fields is correctly reported, it is possible to
calculate farm acreage for a specific crop. However, since the
unique location identifier is made up of three components, errors
or changes in any of these three elements results in double-reporting
of fields. For example, if grower 5003309, growing almonds (site
code 3001) on 20 acres, reports a site locator ID of "A 1"
in one pesticide use report and "A-1" in a second pesticide
use report, this appears as two different locations:
Location
#1: 5003309 3001 A 1
Location
#2: 5003309 3001 A-1
In similar fashion,
errors or changes in the grower ID or site code result in double-reporting
of locations. Examining the three location ID elements, it is possible
to identify some potential problems and solutions.
- Grower
ID Errors
The raw
Grower ID number is a nine-digit number that has two parts.
The first part is a two-digit code for the county in which the
pesticide application report was filed. The second part is a
unique seven-digit ID. Since a grower might file pesticide reports
in different counties, we strip off the first two digits and
use just the unique part of the grower ID. A Grower ID for a
location can change either because of a data entry error or
because the field is sold or leased to a new grower. Errors
and changes in grower IDs are difficult to detect because there
are currently no independent data sources for comparison. In
addition, grower IDs are not formatted for error detection.
(An example of an identification number containing error detection
information is the CAS number, which contains a checksum.
Incorrect CAS numbers can be easily identified by running a
simple math routine). While grower IDs cannot be compared to
independent data, they can be compared to data from a previous
year. Grower IDs that only occur once are likely to be data
entry errors. In a single year, the PUR data contain records
from approximately 28,000 growers. For the 8-year period from
1991 through 1998, the PUR data contain 61,357 unique Grower
IDs. Of these, 4,958 have only one entry in the PUR dataset.
Additional
problems with using Grower IDs to calculate location IDs occur
when a farmer contracts with multiple pest control operators
(PCOs). If each PCO files the pesticide application under their
own ID number it will lead to double-counting of fields. In
addition, counties occasionally assign two different grower
IDs to the same grower.
The sale
or lease of a field can also lead to double counting of fields.
For example, if a 55–acre field of grapes is listed under
one grower ID in a selected MTRS block (the MTRS block is a
geographic identifier – one MTRS block is approximately
one square mile) for several years, and that grower ID disappears
at the same time a second field of 55 acres of grapes appears
under a new grower ID, it is a reasonable assumption that the
field has been sold. In the year that the field was sold, it
will appear twice in the PUR data, once under the old owner
and a second time under the new owner.
- Site
Code Errors
Errors in site codes are less common than other errors because
there are only a limited number of site codes in use, typically
220 in a given year. The small number makes it difficult to accidentally
enter an incorrect value. A notable exception to this is the incorrect
use of a site code to identify the subtype of a crop. For example,
there are two site codes for tomatoes in the PUR data, one for
fresh market tomatoes and a second for processing tomatoes. In
the early years of PUR reporting, the distinction between these
two types of tomatoes was not clear, and many growers used incorrect
site codes.
-
Site
Locator ID Errors
Site
locator IDs are the most unreliable part of the location identifier
in the PUR data. Because site locator IDs are not consistently
formatted, it is very common for site locator IDs to be entered
in several ways. One solution is to strip out common mistakes.
In our data we remove all spaces, dashes, and hyphens in the
site locator ID field. In addition, we convert all G's to 6's,
Z's to 2's, and D's to 0's. This data cleanup technique was
developed by Professor Lynn Epstein at UC Davis.
Top of page
Using
the Location ID to Calculate Crop Acreage
Once grower
IDs, site codes, and the site locator IDs are corrected, these
three fields can be concatenated to create a Location ID. Each
year, there are approximately 180,000 unique locations reported
in the PUR data. One way to calculate total acreage by crop would
be to group the data by Location ID, remove duplicates, then sum
the Acres Planted' field.
While this
technique works, it does not address errors in the Acres Planted
data field. If the Acres Planted are reported correctly, the Acres
Planted values in all records of pesticide use for that location
should agree. In practice, about 96% of all locations have a consistent
Acres planted value for all records in that location set. To determine
Acres Planted for the remaining locations, we calculate the mode
(the most common value) of Acres Planted for the records in the
location set and use this value for all acreage calculations.
This technique does not fix the problem of locations with a single,
very large reported value of Acres Planted. This type of error
has the potential to greatly affect results. For example, in one
year, a single 81,497 acre broccoli field was reported when a
data entry person inadvertently entered the date (8/14/97) into
the Acres Planted data field.
Top of page
Problem
Crops
Calculating
acreage using the techniques described above
works for crops that have one crop per year and are planted and
harvested in the same year. When multiple crops of the same commodity
are grown in one year, it appears as if just a single crop was
grown. As a result, this technique underreports acreage. For crops
with a growing season that spans two calendar years, calculating
planted acreage using this technique can result in up to a twofold
error in reported acreage. Both of these problems can only be
addressed by knowing the planting and harvest dates for each crop,
data that are not currently collected under the PUR system. Fortunately,
most crops are grown once per season and are planted and harvested
in the same year. Some notable exceptions are carrots, garlic,
cauliflower, celery, cabbage, lettuce (head and leaf), broccoli,
onions, asparagus, strawberries, and spinach. In the PAN search
results, a comment is inserted for these crops to alert the user
to potential problems with the data.
Top
of page
Evaluation
of the Accuracy of Acres Planted
In order
to evaluate the accuracy of the Acres Planted determined from
the PUR data, we took the
following steps:
-
PUR
Acres Planted is compared across all years for which data are
available. Any extreme changes in acreage are flagged for further
investigation.
- Any known
crop trends are evaluated in the context of the data. For example,
it is widely known that almond acreage increased in California
during the 1990s. Another example is the failure of a fraction
of the cotton crop due to the extremely wet and cold El Niño
year in 1998, resulting in a decrease in acres planted because
approximately 100,000 acres of cotton was either not planted or
failed due to the weather before any pesticides were applied.
- To determine
whether an unexpected change in Acres Planted might be due to
an actual change in acres planted or was more likely to be an
error in the data, three other PUR data fields are evaluated,
the pounds of pesticide applied, the number of locations treated
and the number of applications made. For example, if all of these
parameters increased concurrently with the number of acres planted,
the increase in acres planted is likely to be real.
- The change
in acres planted is compared to the range of what is realistically
possible for that particular crop. For orchard and vine crops
that take several years to reach maturity and are viable for many
years of harvests, any rapid swings in acres planted from year
to year is almost certainly an error in the data. For annual crops,
a change in commodity prices could easily influence how many acres
are planted in that particular year.
- Because
of the way the PUR data are collected, PUR acreage estimates are
never reliable for certain problem crops.
-
Finally,
PUR Acres Planted is compared to the California Department of
Food and Agriculture (CDFA) Acres Harvested, when these data
are available. However, one must view the CDFA data with some
skepticism. There are several reasons these data are not necessarily
accurate:
- CDFA estimates
routinely underreport crop acreage because they are based
on grower surveys. Typically only 80% of growers who receive
the surveys return them, and the surveys are not sent to all
growers of the crop in question. (For example, see CDFA's
comment on incompleteness of their acreage estimates made
in the 1998
California Almond Acreage Report, California Department
of Food and Agriculture, May 25, 1999). Thus, in most cases
the PUR data provide a more accurate estimate of acres planted
than the CDFA data.
- CDFA estimates
are for acres harvested. Since pesticides are also applied
to crops like young trees or vines that are not necessarily
harvested that year, this estimate does not capture the entire
picture of acres planted for that crop.
Last updated
November 11, 2002
|
|
