- PUDS Methodology and Data Sources
- Dealing With Data Gaps
- Dealing With Missing Years
- Further Issues
Access the Pesticide Use Data System (PUDS) HERE
- Pesticides Introduction
- Pesticide Usage
- Dietary Risks of Pesticides
- The Dietary Risk Index
- Impacts of GE on Pesticide Use
- Environmental and Other Impacts of Pesticides
PUDS provides access to detailed information on pesticide use by crop and year, at the national and state level. The system draws upon pesticide use data collected and published annually by the USDA’s National Agricultural Statistics Service (NASS), via their online Quick Stats Database.
NASS conducts farm-level surveys of pesticide use. Major row crops are surveyed in most years, while fruit and vegetable crops are surveyed bi-annually (fruit crops in odd years, vegetables in even years.
In each year for all crops surveyed, NASS collects data in states that collectively account for at least 85% of the total acres planted nationally to a given crop.
PUDS Methodology and Data Sources
PUDS data elements are downloaded from two separate components of Quick Stats. Total acres planted to a specific crop are reported by state and national totals, and are extracted from the Quick Stats crops sector. All other pesticide use data is reported under the environmental sector.
Key PUDS data elements include:
Crop – Crop surveyed by NASS.
Year – Year the crop was surveyed for pesticide use.
State – The state the crop was planted in. ‘National’ designates national use of the pesticide.
PC Code – A code for each registered pesticide active ingredient that is assigned by the EPA, and used to link to all EPA data on pesticide toxicity and environmental fate and impacts.
Active Ingredient Type (AI Type) – Active ingredient type: herbicide, insecticide, fungicide, or other type of pesticide. Occasionally, a pesticide may be used as an herbicide in one treatment and a fungicide in another treatment.
Chemical – The pesticides applied to crops are reported according to their PC code and active ingredient name as designated by the EPA. Use of brand named products (i.e. Roundup (glyphosate) or Admire (imidacloprid)) are not surveyed nor reported in Quick Stats.
Conglomerate – NASS sometimes reports pesticide use separately for different chemical forms of the same active ingredient (e.g., 2,4-D diethamine salt and 2,4-D isoprop ester). The conglomerate methodology in PUDS groups all such related compounds into an aggregate total for the active ingredient (see below for details).
Total Acres – The total number of acres that is planted to a specific crop for each state, and national totals. “Total Acres” at the national level typically exceeds the sum of NASS surveyed acres.
Note — In the case of some crops (fruit and nut trees, bushes and vines), USDA may report bearing acres as opposed to planted acres, since these crops are not replanted annually. In such cases, acres bearing, or acres harvested in a specific year is included in PUDS as Total Acres Planted value.
Surveyed Acres – The number of acres planted, as surveyed and reported by NASS within each state.
Note — Surveyed acres is not the same as total acres planted, for two reasons. First, USDA usually surveys up to 85% of the acres of a given crop in each state surveyed. Second, not all states growing a given crop are surveyed.
Accordingly, PUDS use tables report values for both “Surveyed Acres” and “Total Acres.” Total pesticide use at the state and national levels are estimated by assuming that the percent acres treated and rates per crop year are the same on the acres not surveyed by NASS as on the acres surveyed.
Percent Acres Treated – The percentage of total crop acres within a state that was treated with a given pesticide. In Quick Stats, this data item is labeled Treated, measured in pct of area planted, avg.
Number of Applications – The average number of times a pesticide was applied within a crop year. The Quick Stats data item is Applications, measured in number, avg.
Rate of Application – The average one-time rate of application per acre for the pesticide active ingredient. The Quick Stats data item is Applications, measured in lb / acre / application, avg.
Rate per Crop Year – The average one-time rate of application per acre multiplied by the average number of applications in a crop year. The Quick Stats data item is Applications, measured in lb / acre / year, avg.
Pounds Applied – The pounds of pesticide active ingredient applied to the treated survey acres in a state, or nationally. The Quick Stats data item is Applications, measured in lb.
Calculated Pounds Applied to Surveyed Acres – The pounds of pesticides applied to surveyed acres as reported in Quick Stats data.
Calculated Pounds Applied to Total Acres – The pounds of pesticides applied to total acres planted of a given state/crop.
Surveyed Year – Crop year for which pesticide use data were cllected.
Dealing with Data Gaps
There are instances in Quick Stats where a data element is reported as undisclosed (D), or as insignificant data (Z). In such cases, there may only be one farmer within a state growing a specific crop, which renders the data unreportable, because of NASS privacy protection policies.
PUDS includes a method to estimate most unreported values from those data elements that are reported.
NASS sometimes reports the use of an active ingredient in two or more, slightly different chemical forms. PUDS has the ability to group multiple forms of an active ingredient together into what we call “conglomerate” value.
The application rates and number of applications for conglomerate chemicals are calculated using a weighted average approach, based on the percent of acres treated with each individual form of the active ingredient, relative to the sum of acres treated to all forms. Conglomerate pounds applied is simply the sum of all chemical variations.
Selected PUDS tables will be accessible in two forms, one reporting each form of a given active ingredient separately, and the other reporting conglomerate values.
Dealing with Missing Years
The U.S. Department of Agriculture has collected reasonably comprehensive pesticide use data for major grain, row crop, fruit, and vegetable crops since 1990. A few additional surveys collected use data on some major crops going back into the 1970s. PUDS contains the results of all pesticide use surveys conducted by the USDA.
Because of budget limitations in the last decade, NASS has limited the number of crops surveyed in any given year. For example, soybean pesticide use was surveyed annually from 1990-2002, in 2004-2006, and then not again until 2012 and 2015. Pesticide use on corn was surveyed each year from 1991-2003, in 2005, 2010, 2014 and 2016.
The absence of annual survey data for many crops creates gaps in PUDS. To overcome this shortcoming, a series of methods have been developed to approximate missing data values by interpolating between years with reported values. The default assumption is that pesticide use changes by the same proportional amount year to year between two years with known values.
Missing data values are estimated as follows:
- Years prior to the first year an AI had reported usage, the data values are set to zero.
- Years after the last year an AI had reported usage, the data values are extrapolated forward as equal to the last surveyed year.
- If an AI did not have usage data reported in a given year where a crop/state was surveyed, then it is set to zero. (See below for further implications).
- For years between a surveyed value and a zero value, it is assumed that use declines linearly. The percent acres treated are interpolated from the surveyed value to zero. Meanwhile, the rate of application, number of applications, and rate per crop year, remain the same as the previous/latter year.
- All other years are interpolated between two surveyed years.
There are cases where Quick Stats does not fully report use data for some pesticide-crop combinations when: (a) the pesticide is applied on less than 1% of acres, (b) a trivial amount of a low-dose active ingredient is applied, or (c) when only one or a few producers reported use of the pesticide, raising confidentiality concerns.
In such cases, use of the unreported AI’s are estimated by subtracting the sum of pounds applied for reported AI’s from the total pounds applied as reported by Quick Stats for each AI type (H/I/F/O)).