Survey Results Part 1

Jump to another section
Go to Section 2 – Crops and Pesticide-Related Bee Kills –>
Go to Section 3 – Economic Impacts and Bee-Kill Investigations –>
Go to Section 4 – Additional Comments from Question 15 –>
Go to Section 5 – Description of Methods Used –>

Why a Survey on Acute Pesticide-Related Bee Kills?

This survey on pesticide-related bee kills was developed in service to the US EPA Pesticide Program Dialog Committee (PPDC) Pollinator Workgroup, with the intent of determining if there are particular crops that are more or less problematic for acute pesticide poisonings of bees. Most of the questions focus on acute bee kills caused by high doses of pesticides, including insecticides, fungicides and herbicides. While sub-lethal pesticide effects may also be relevant to the issue of declining bee populations, these are beyond the scope of this survey, as multiple contributing factors may be involved given the time between exposure and the observation of adverse effects. As a result, the focus of this survey was primarily on acute bee kills caused by pesticides, with only a preliminary analysis of the relationship of crops on which bees foraged to long-term effects on bee health.

The survey was designed by Pesticide Research Institute working together with several commercial beekeepers who are members of the PPDC Pollinator Workgroup, including Darren Cox of Cox Honey in Utah, Jeff Anderson of California-Minnesota Honey Farms, Rick Smith of Smith Honey, Bret Adee of Adee Honey, and Steven Coy of Coy’s Honey Farm. Data processing and analysis was done by Josh Pepper, Sandra Bustos, Ted Callon and Susan Kegley of Pesticide Research Institute, with external review provided by the beekeepers listed above, as well as PPDC Pollinator Protection Group members Erik Johansen of the Washington State Department of Agriculture and Iain Kelly of Bayer Crop Science. We hope the results will provide useful data to US EPA.

Message to Beekeepers and Participants

We would like to thank all of the beekeepers who participated in this survey on pesticide-related acute bee kills. Your input was invaluable. We very much appreciate your support and participation.

Key Findings

Below are some highlights from the survey results. For the methods used to do the data analysis, see the Methodology section.

Survey Respondents

  • Both commercial and non-commercial beekeepers filled out the survey. There were more non-commercial respondents, but the majority of the hives were owned by commercial beekeepers. Respondents were categorized as either non-commercial or commercial beekeepers based on the number of hives they manage, with 50 hives set to be the breakpoint to distinguish commercial versus non-commercial beekeepers. Out of the 427 individuals who completed the survey, 365 were categorized as non-commercial beekeepers, who altogether accounted for 2,597 unique queen-right colonies, hereafter referred to as hives. Sixty-two respondents were categorized as commercial beekeepers, accounting for 244,171 unique hives.
  • The total number of hives that spent time in any state was estimated at approximately 900,000. This method counts hives once for every state they spent time in, like the National Agricultural Statistics Service’s (NASS) hive counts and assumes that, on average, only 75% of a beekeeper’s total number of hives foraged in each state. See the Methodology section for more details and caveats on comparing different methods of estimating hive numbers.

General Observations

  • Commercial beekeepers reported acute pesticide-related bee kills more frequently and observed a higher number of acute bee kills and longer-term abnormal bee mortality not clearly related to acute poisoning events than non-commercial beekeepers. Sixty-eight percent of non-commercial beekeepers reported never having experienced a pesticide-related kill in the last three years, while only 12% of commercial beekeepers reported not having a kill.
  • When asked what factors they believe contributed to acute pesticide-related bee kills, 76% of commercial beekeepers cited pesticide use on blooming crops not being commercially pollinated by their bees and 74% indicated pesticide use on crops not dependent on pollination (such as corn) as the top two factors. In contrast, the top two factors cited by non-commercial beekeepers were pesticide use on non-agricultural land (e.g. golf courses, landscaped areas, and roadside weeds), which was chosen by 29% of non-commercial respondents, and pesticide use on crops not dependent on pollination, which was chosen by 24% of non-commercial respondents.

Analysis of Bee Kills by Crop

Click here to go directly to the detailed results by crop.

The crop analysis was also split out by commercial and non-commercial beekeepers, and the analysis was done three different ways: 1) By the number of respondents experiencing kills occasionally or frequently, which provides information on the number of beekeepers with acute poisoning experiences by crop; 2) By the number of hives experiencing kills occasionally or frequently, which provides information on which crops affect the largest number of hives; and 3) By the number of hives affected by a particular crop relative to the acres of crop planted in the U.S., which provides information about the relative per-acre hazard each crop poses to bees.

By Number of Respondents:

  • Top crops cited by commercial beekeepers for acute pesticide-related bee kills were cotton (71%), corn (54%) and melons (40%).
  • The top crops cited by non-commercial beekeepers for acute pesticide-related bee kills  were corn (21%), soybeans (17%) and urban plants (14%).

By Number of Hives:

The analysis by number of hives was only done for commercial beekeepers. The top crops cited by commercial beekeepers for acute pesticide-related bee kills analyzed by number of hives were cranberries (94%), alfalfa hay (91%) and soybeans (87%).

By Number of Respondents and Number of Hives per Acre of Crop Planted:

We defined a respondent Bee-Kill Index (rBKI) and a hive Bee-Kill Index (hBKI) to express the relationship between hives experiencing kills and the crop acreage planted in the U.S., with higher values of the BKI indicating a higher frequency of acute kills per acre of crop planted than for lower values of the BKI.

Respondents

The analysis by number of respondents per acre of crop planted was only done for commercial beekeepers and was broken out into two categories: 1) Commercially pollinated crops, and 2) Crops not utilizing commercial pollination. The commercially pollinated crops with the highest rBKI were winter squash (rBKI = 6.9), summer squash (rBKI = 6.7) and melons (rBKI = 4.6). All of these crops bloom throughout the season, increasing the potential for pesticide exposure.

For crops not utilizing commercial pollination, cotton (rBKI = 0.048) was the crop of greatest concern, followed by alfalfa (rBKI = 0.016), corn (rBKI = 0.006), and soybeans (rBKI = 0.005). Although cotton has the smallest number of acres planted in this category of crops at 14.7 million acres, it was the most problematic for bee kills, indicating that pesticide applications to this crop are responsible for a large number of observed bee kills per acre of crop planted. In spite of the relatively low rBKI for soybeans and corn, the large number of acres of the crop planted results in high exposure potential for bees in the areas where these crops are grown, with 54% and 37% of commercial respondents reporting kills occasionally or frequently on corn and soybeans, respectively.

Hives

The analysis by number of hives per acre of crop planted was only done for commercial beekeepers and was separated into commercially pollinated crops and crops not utilizing commercial pollination. The data are shown in terms of the hive bee-kill index, hBKI. The top commercially pollinated crops cited by commercial beekeepers for acute pesticide-related bee kills per acre planted were cranberries (hBKI = 24), winter squash (hBKI = 13) and melons (hBKI = 8.0).

For crops not utilizing commercial pollination, alfalfa hay (hBKI = 0.043) and cotton (hBKI = 0.039) were the crops of greatest concern. The hBKI for soybeans and corn was approximately one quarter of the values for alfalfa and cotton. This result may be due to the low numbers of commercial respondents from corn and soybean-growing areas and the generally low number of commercial beekeepers that keep their bees in these areas. While corn and soybeans had low numbers of hives affected per million acres, the absolute number of hives with kills attributed to corn and soy was still high, with 74% (corn) and 82% (soybeans) of the hives reported as experiencing kills on these crops occasionally or frequently.

The great disparity between commercially pollinated crops and those not requiring commercial pollination makes some sense, with a primary contributing factor being the number of acres planted. Other considerations include:

  1. Use of bee-toxic pesticides varies by crop depending on pest pressure and types of pests.
  2. Bees are more likely to be located near commercially pollinated crops than crops not requiring pollination.
  3. Most of the commercially pollinated crops with high rBKI or hBKI values have a relatively long bloom period: Winter squash, melons, cucumbers, alfalfa seed, and summer squash, which means there are more opportunities for acute pesticide poisonings. Cranberries are an exception; with only a 1–2 week bloom period, this reasoning cannot explain the result. However, results for cranberries are based on the response of a single beekeeper, and the information for questions 6 and 7 is conflicting for this respondent.
  4. Most of the commercially pollinated crops with low rBKI are crops for which commercial pollination services are not always utilized or to which fewer pesticides are applied (sunflowers, canola) or for which well-defined bloom periods allow for removal of bees prior to pesticide applications (almonds, cherries, citrus, apples and pears).
  5. Beekeepers reported attempting to stay away from certain problematic crops, such as cotton and alfalfa (see question 8). In spite of this fact, both of these crops had relatively high BKI values compared to corn and soybeans, which have many more acres planted.
  6. Bees may be more or less attracted to different crops.
  7. Different regions of the country were not equally represented in the survey, with neither the corn/soybean-growing states nor the cotton-growing states fully represented. As a result, the relative ranking of the non-commercially pollinated crops indicated by the BKI values is suggestive, but not conclusive.

Longer-Term Hive Dwindling and Loss

When asked what percentage of their hives exhibited symptoms of abnormal bee mortality over the past three years, 58% of commercial beekeeper respondents said that at least 20% of their hives had exhibited such symptoms. Only 37% of non-commercial beekeepers said that at least 20% of their hives had exhibited such symptoms.

Economic Impacts and Kill Investigations

Click here to go directly to the detailed information on economic impacts and kill investigations.

  • When commercial beekeepers were asked how often they had moved their hives in the past three years to protect their bees from pesticides and if there was any related increase in expenses, 59% said they had moved their hives at least once. Only 12% of non-commercial beekeepers moved their bees to avoid pesticide exposure. Of the commercial beekeepers that moved their hives to avoid pesticide exposure, 97% reported that this has led to “some” or “substantial” additional expense or lost income.
  • Thirty-seven percent of commercial beekeepers that had bee kills did not report them to enforcement personnel. Based on the comments from commercial beekeepers, many are dissatisfied with bee kill investigations based on their past experience with reporting pesticide-related bee kills. Several commercial beekeepers noted that their failure to report was related to their dissatisfaction with bee kill investigations in the past.
  • Twenty-four percent of non-commercial beekeepers that had kills did not report them to enforcement personnel. Several of these beekeepers stated that they did not know how to document or report a bee-kill incident.
  • When asked “If you reported a bee kill to the authorities in your state, what response did you get?” no respondents selected the option: “For most incidents, a violation was issued by the investigator and the pesticide applicator was fined.”

Summary of Other Comments

Click here to go directly to the collection of all other comments.

  • Many respondents said that pesticide labels need to include more precautions about how to apply pesticides without harming bees. Specifically, that labels should include a simple system with symbols such as color-coded bees that indicate the pesticide bee toxicity of a particular product. Other respondents, especially non-commercial beekeepers, suggested that more education on properly adhering to pesticide labels is needed, especially for homeowners.
  • Commercial beekeepers indicated that it is critical that beekeepers, growers, and pesticide applicators communicate in order to effectively coordinate pesticide application times and reduce bee kills.
  • Many beekeepers expressed their concerns about herbicide use on crops and non-agricultural land along with insecticide use as specifically problematic for bees.
  • Several beekeepers mentioned that growers and applicators are not consistently adhering to local ordinances that require that beekeepers be notified before pesticides are applied near their hives or that in some cases prohibit commercial applicators from spraying crops within a certain radius of registered hives.
  • Beekeepers also indicated that acute bee kills are not the only issue related to pesticide use, and expressed their concerns that exposure to pesticides at low doses over time may also be responsible for queen failure and reduced overwintering survival of hives.

Profile of Survey Respondents

Question 5: How many respondents were there from each state?

As of June 6th, a total of 427 beekeepers from all 50 states had responded to the survey. The states with the highest number of respondents included:

  1. California (52 respondents)
  2. North Carolina (38 respondents)
  3. Ohio (30 respondents)
  4. Maine (28 respondents)

The number of survey respondents from each state is not necessarily representative of the total number of beekeepers and/or hives in each state because of sampling limitations. See Methodology for more information about limitations of the survey.


Note: In situations where a respondents indicated that their hives had been in more than one state, the respondent was counted more than once. Thus, the total number of respondents represented in this map (487 respondents), is larger than the number of unique respondents reported in the survey (427 unique respondents) because beekeepers that traveled to more than one state are counted once for each state in which their bees foraged. Respondents that did not list any states are not included in this data.

Questions 1 and 5: How many beehives were there in each state?

Beehives from all 50 states, totaling 246,768 unique beehives are represented in the survey results. Out of the 427 individuals who completed the survey, 365 were categorized as non-commercial beekeepers with fewer than 50 hives, who altogether accounted for 2,597 unique queen-right colonies, hereafter referred to as hives. Sixty-two respondents were categorized as commercial beekeepers with more than 50 hives, accounting for 244,171 hives. Beehives were counted once for every state they spent time in, like the National Agricultural Statistics Service’s (NASS) hive counts. For the purposes of the crop analysis (question 7), we assumed that only 75% of the total hives reported in the survey were pollinating each crop. Using these two assumptions, we estimate that the total number of hives that spent time in any state (including multiple counting of unique hives) is approximately 900,000 beehives. See Methodology for this question.

The states with the largest number of hives were:

  1. California (223,986 hives)
  2. North Dakota (143,600 hives)
  3. South Dakota (106,401 hives)
  4. Texas (105,039 hives)
  5. Maine (103,715 hives)


Note: In situations where respondents indicated that their hives had been in more than one state, the number of hives was counted more than once. Thus, the total number of hives represented in this map (1,214,134 total hives), is larger than the number of unique hives reported in the survey (246,768 unique beehives) because hives that traveled to more than one state are counted once for each state in which they foraged. Respondents that did not list any states are not included in this data.

Question 2: Is your beekeeping operation for honey production, commercial pollination services, or both? 

Click here for description of results

When asked to describe the extent of their beekeeping operations, 61% of respondents (consisting of 0.5% of the total number of hives represented in the survey) replied that they practice hobby beekeeping, 22% of respondents (consisting of 3% of the total number of hives represented in the survey) said they focus only on honey production, 11% of respondents (consisting of 65% of the total number of hives represented in the survey) said they focus mainly on honey production but provide some commercial pollination services, and 6% of respondents (consisting of 28% of the total number of hives represented in the survey) said they focus mainly on commercial pollination services but do some honey production as well. See Methodology for this question.

Question 3: If you do provide commercial pollination services, what percent of your hives, on average, is used for these services?

Click here for description of results

When respondents who do provide commercial pollination services were asked about the percentage of their hives that they use for such services, 41% said they use 75-100% of their hives, 26% said they use 0-25% of their hives, 24% said they use 25-50% of their hives, and 9% said they use 50-75% of their hives.

A total of 46 commercial beekeepers with 225,916 hives responded to this question.

See methodology for question 3

Continue to Section 2 – Crops and Pesticide-Related Bee Kills –>

UA-21813186-1