Reverse a ban? Review the science

lower klamath refuge USFWS 960x
Lower Klamath National Wildlife Refuge in California and Oregon is one of several refuges that could now begin using genetically modified crops and neonicotinoid insecticides. / Photo source: U.S. Fish and Wildlife Service

Early this month, the Trump administration ended restrictions on the use of neonicotinoid insecticides in national wildlife refuges. The change was announced in a memo from the U.S. Fish and Wildlife Service that was primarily intended to rescind the ban on the use of genetically modified crops within the refuge system. “In addition, I am withdrawing the 2014 memorandum’s restrictions with regard to neonicotinoid pesticides that are often used in conjunction with GMO seed,” wrote Gregory J. Sheehan, Principal Deputy-Director of USFWS, as one of his final actions before leaving the wildlife service this week.

The 2014 memo to which he referred was released during the Obama administration after a decision by the National Wildlife Refuge System’s Leadership Team that only agricultural practices which specifically contribute to wildlife objectives would be allowed in the refuge system. As part of that decision, the Leadership Team stated that neonicotinoid pesticides would no longer be permitted: “We have determined that prophylactic use, such as a seed treatment, of the neonicotinoid pesticides that can distribute systematically in a plant and can potentially affect a broad spectrum of non-target species is not consistent with Service policy.”

The reversal of the 2014 ban is just one in a series of federal actions focused on changing how pesticides are used and regulated in the United States. This includes the recently-proposed changes to the Endangered Species Act, a memorandum of agreement between the Environmental Protection Agency, Interior and Commerce departments, and certain sections in the House version of the 2018 Farm Bill.

The stated reason for ending the ban on neonicotinoids was that these pesticides “are often used in conjunction with GMO seed”, and farming practices that employ genetically modified crops could be necessary “to best fulfill the purposes of the refuge and the needs of birds and other wildlife” as described in other sections of the memo.

However, Sheehan did not address the original concern that these same pesticides “can potentially affect a broad spectrum of non-target species” including bees, nor did he explain how such collateral impact would be consistent with any specific objective.

The extent to which the new policy could or will affect bee communities in North America isn’t clear yet. It will obviously depend on what sorts of bees live in and around these various refuges. It will also depend on how refuge managers choose to employ this new flexibility and how rigorously they adhere to the wildlife service’s own recommended best practices regarding pesticide use.

What is clear, however, is our scientific knowledge on the matter.  The same week this change was announced, the California Department of Pesticide Regulation released a 1100-page report laying out how neonicotinoids can be harmful to bees and other pollinators. So let’s use this as an opportunity to begin reviewing what we know.

Below is a new, on-going project from the Bee Report: a review of the lethal and sub-lethal effects that this class of insecticides can have on bees. The review is currently based on 49 different studies. A list of 96 different effects, grouped into nine different categories, has been created from this review. A quick look at the list shows us that the majority of effects fall into the categories of reproduction, foraging and physical behaviors.

effects chart COMPLETEB

To dive even deeper into the review, click on a category below for drop-down information. You’ll see the specific effects, the types of bees that showed those effects, the types of neonics that caused the effects and the research references for that information. You can open multiple categories at the same time. Just be warned, the list is long and can feel a bit overwhelming. Which is actually the point: the evidence that neonicotinoids can be harmful to bees is overwhelming.

This project is a work in progress and will continue to be updated. If you have additional research that you think should be included, please get in touch with me at

EffectType of ChangeType of BeesType of NeonicsReferences
DecreaseHoney bees Acetamiprid (AC), Clothianidin (CL), Dinotefuran (DN), Imidacloprid (IM), Thiamethoxam (TH), Thiacloprid (TC)

"&" indicates a mixture
IncreaseBumble Bees
AlteredSolitary Bees
bee immobilityIMMoffat et al. 2016
breathingIMHatjina et al. 2013
feedingTHElston et al. 2013
hygienic behaviorIMTsvetkov et al. 2017; Wu-Smart & Spivak 2016
locomotor skills in queensIMWu-Smart & Spivak 2016
motor functionCL, DN, IMWilliamson et al. 2014
pollination servicesTHStanley et al. 2015b
proboscis extensionAC, THAlkassab & Kirchner 2016; Démares et al. 2016; Thany et al. 2015
swarmingCL&THSandrock et al. 2014a
time groomingTHWilliamson et al. 2014
worker movementCL, IM, THScholer & Krischik 2014
learningTHStanley et al. 2015a
memory, long-term CL, IMAlkassab & Kirchner 2016; Williamson & Wright 2013
memory, short-term and mid-term IM, THStanley et al. 2015a; Williamson & Wright 2013
foraging activityCLArce et al. 2016; Wu-Smart et al. 2016
flight duration (chronic exposure)THTosi et al. 2017
flight duration (acute exposure)THTosi et al. 2017
flight distance (chronic exposure)THTosi et al. 2017
flight distance (acute exposure)THTosi et al. 2017
flight velocityTHTosi et al. 2017
forager recruitmentIMGill et al. 2012
foraging performanceIMCresswell 2011
foragers returning to patchIMKarahan et al. 2015
foraging tripsIMKarahan et al. 2015
homing capacityCL, IM, THFischer et al. 2014; Yang et al. 2012
pollen loadIMGill et al. 2012; Stanley et al. 2016
pollen storesIMWu-Smart & Spivak 2016
pollen trip (duration)IMGill et al. 2012
pollen trip (successful)IMGill et al. 2012
rate of return to colonyTHStanley et al. 2016
time spent foragingTHStanley et al. 2016
workers lost during foragingIMGill et al. 2012
daily mortalityIMAbbo et al. 2017
life spanCLStraub et al. 2016; Tsvetkov et al. 2017
mortalityIMAlaux et al. 2010; Trayner et al. 2016
mortality (synergistic effects)CL, THSgolastra et al. 2016; Zhu et al. 2017
worker mortalityIM, THMommaerts et al. 2010
parasites and pathogens (abundance)variousSanchez-Bayo et al. 2016
parasites and pathogens (spread)variousSanchez-Bayo et al. 2016
immune responseIMCzerwinski & Sadd 2017
varroa infestationCL&THAlburaki et al. 2015; Alburaki et al. 2018
body massIMAbbo et al. 2017
depolarization in neuronsIMMoffat et al. 2015
gene expression (variety of genes)AC, CL, IM, THChristen et al. 2016; Simmons & Angelini 2017
hemocyte countCLBrandt et al. 2016; Hernandez-Lopez et al. 2017
neural stimulationCL, IMMoffat et al. 2016
neural sensitivity to substancesIMMoffat et al. 2015
size of hypopharyngeal glandsIMAlaux et al. 2010; Hatjina et al. 2013
thermoregulationTHTosi et al. 2016
vitellogenin levelsIMAbbo et al. 2017
wild bee densityCLRundlöf et al. 2015
reproductionCLRundlöf et al. 2015
reproductive successCL&THWilliams et al. 2015
adult workers, drones, gynes numbersCLArce et al. 2016
brood cellsIM, TH, CL&THMoffat et al. 2016; Sandrock et al. 2014b
brood developmentIMGill et al. 2012
brood productionIMWu-Smart & Spivak 2016
brood sizeCL&THSandrock et al. 2014a
colony failureTCEllis et al. 2017
colony growthCL, IMBryden et al. 2013; Rundlöf et al. 2015
colony survivalIMTasei et al. 2000
colony weightTCEllis et al. 2017
completed nestsCL&THSandrock et al. 2014b
dead coloniesCL&THAlburaki et al. 2015; Alburaki et al. 2018
dronesIMMommaerts et al. 2010; Woodcock et al. 2017
drone productionIM, THMommaerts et al. 2010
egg cellsCL, THWoodcock et al. 2017
eggs laidIM, THElston et al. 2013; Wu-Smart & Spivak 2016
femalesTHMoffat et al. 2016
females (proportion to offspring)CL&THSandrock et al. 2014b
larva producedTHElston et al. 2013
larval moralityCLHernandez-Lopez et al. 2017
likeliness to loose queenAC, CL, IM, THTsvetkov et al. 2017
nest buildingTHElston et al. 2013
nest reproductionIMMommaerts et al. 2010
offspring able to hatchCL&THSandrock et al. 2014b
offspring completing developmentCL&THSandrock et al. 2014b
offspring productionCL&THSandrock et al. 2014b
ovary sizeCL&THWilliams et al. 2015
population size of adultsCL&THSandrock et al. 2014a
queens laying eggs (percentage)CL&THWilliams et al. 2015
queen productionCL, IM, TH, IM&THMoffat et al. 2016; Whitehorn et al. 2012; Woodcock et al. 2017
queen replacementCL&THSandrock et al. 2014a
queen survivalCL, IMFauser et al 2017; Scholer & Krischik 2014
reproductive cellsCL, IM&THWoodcock et al. 2017
reproductive individualsTCEllis et al. 2017
solitary bee nestingCLRundlöf et al. 2015
sperm qualityCL&THWilliams et al. 2015
sperm viabilityCL, THStraub et al. 2016
storage cellsCL, THWoodcock et al. 2017
time of developmentCL&THAbbot et al. 2008
wax cells builtTHElston et al. 2013
worker fecundityIMLaycock et al. 2012
workersCLWoodcock et al. 2017
workers producedTHStanley et al. 2016
olfactory associative behaviorIMYang et al. 2012


This review is based predominately on two sources of information:

1. Supplementary materials from Mitchell and colleagues’ 2017 study, “A worldwide survey of neonicotinoids in honey”. Table S8 in those materials contains an extensive list of effects and references to the studies demonstrating each effect.

2. The IPI database which contains summaries of research articles on pesticides and their effects on invertebrates. The articles in the database have been reviewed and summarized by Xerces Society staff. This list was built from the results of a search for “neonicotinoids, bees”.

Effect: The impact descriptions were taken from the summaries of research in the supplemental materials and the IPI database. When necessary, I referred to the abstracts of the original research work for clarification. In listing the effects and grouping together different studies with similar effects, I was particularly careful not to combine effects that could be subtly but distinctly different (ex. decreases in “nest building” and “nest reproduction” are not necessarily the same thing); I preferred to risk duplicating effects on the list than to miss an important distinction between them.

In the interest of making the list easier to navigate and read, I grouped together effects into different categories that seemed to make sense.

Type of Change: The direction of each impact – “decrease” or “increase” – was also taken from the summaries of research in the supplemental materials and IPI database. When both “decrease” and “increase” are listed, it indicates mixed results within a single study or between studies. “Altered” is my own category for results that did not seem best represented by “decrease” or “increase” or both.

Type of Bees: The category “Honey bees” includes Apis mellifera. “Bumble bees” includes Bombus terrestris and Bombus impatiens. “Solitary bees” includes Osmia bicornis with a single occurrence of Osmia lignaria.

This project is on-going. If you have additional research that you believe should be included in the review, please get in touch with me at