Bees and neonicotinoids, and choice of database

There are petitions circulating online at the moment about the effect of neonicotinoid insecticides on pollinators like bees, and asking for those insecticides to be banned.

Neonicotinoid insecticides were banned, it is said on scientific advice, but now are not, it is said on pressure from the manufacturers and farming interests. 

I have been signing the petitions, having become interested in bees since we had tree bumblebees living in the roof of our last house, and as part of my general pottering in the garden.  I have been involved in research into them and other pollinators by taking part in the University of Sussex's Bees 'N' Beans project (beans now harvested, keeping the results safe until I can submit them online!).   

Until we know the effects on pollinators for sure, we ought not to use neonicotionids. 

What is the evidence?

I searched PubMed for the two words bees and neonicotinoids, and as a test of its usefulness in life sciences more generally.  Limiting to review articles (although we do need to remember the potential pitfalls of review articles) finds this from 2014 by seven researchers, all at the time UK based except one, based in the Netherlands.  It summarises the evidence in several appendices and discusses laboratory experiments (where the dosage may be higher than in the field) and field experiments.

The search finds 69 items.  The search term bees is mapped to MeSH, but the other term is not.  So, I tried truncating neonicotinoid* and this finds 110.  Truncating in PubMed turns off the MeSH mapping, so is not always wise - you may find more with your search term in the title, but lose items found with MeSH mapping.  Here, though, since the first term did not map to MeSH, truncation is wiser.  

A search of Web of Science Core Collection for 

(bee or bees) and neonicotinoid* 

finds 234.  13 of these are reviews.

Importing the 110 and the 234 into RefWorks and removing close duplicates (I have to confess I did not check them all, I assumed they were indeed duplicates) leaves 157 (out of 234).    More work would be needed to see how many of those were unique to PubMed or to Web of Science. 

A general web search finds a lot of news and campaign information, but also this research review from the Xerces Society for Invertebrate Conservation, on the website of the Entomology department of Penn State University.