Too simple reasoning.
- Do legless people go walking when you stop cutting legs?
A bad hive is a bad hive for years.

- "can you show me the study where" bees nourished with neonicotinoids are unarmed? I mean not written by the producers or sellers indeed.

- Where did you find that CCD is a one factor problem. We also have varoa and virus and now Asian hornet!

Find a way to kill Asian hornet with neonicotinoids and I vote for it.

You see it is not as simple as neonicotinoids=no bees, no neonicotinoids=happy bees.

For me this year is the best I had in the 10 last years. Few Asian hornet, no dead colony and a reasonable crop.

Sylvain, sure, declining bee colonies is a very serious, complex, and severe problem. However, if I remember correctly, Dr. Manners referred to a study showing that imid cannot enter flowers if applied to root zone. How can bees be affected in this case??

Also, if I understand correctly, the EU ban was more-or-less triggered by the incident when German (?) farmers (?) dumped large quantities of chemicals (that happened to be neocorts) on a field where bees were foraging. Sure, they were killed (bees, not farmers)! But I bet you a nickel, the same could happen with any "safe-to-use" chemicals too. Perhaps, banning careless people instead of chemicals would make more sense than...

UC Davis seems to disagree with the notion that imidacloprid is not translocated to the flower:
http://www.ipm.ucdavis.edu/TOOLS/PNAI/pnaishow.php?id=42

The Florida Dept. of Ag. has also recognized that imidacloprid and other neonicitinoids appear in nectar, even when applied as a soil drench:
http://www.freshfromflorida.com/content/download/12723/226625/Citrus_Producers.pdf

See also:
Imidacloprid translocated to buckwheat nectar kills wasps: http://www.entomology.umn.edu/cues/krischiklab/pubs/Krischik_et_al._2007.pdf

Imidacloprid is translocated to sunflower pollen in amounts of ~3.4ug/kg. An average colony (20,000 bees) brings in around 60 kg of pollen annually, so 3.4 x 60 = 204 ug; LD50 for bees is between 4-71 ng, and effects on bee behavior are observed in the field beginning at 0.25 ng (0.07 ng in tunnel studies). Thus, an average colony bringing in pollen only from imidacloprid treated plants (not actually far fetched for agrobusiness bees - it's actually the goal to restrict bee foraging to crops) may be, over the course of the foraging season, bringing in enough imidacloprid to kill itself twice over, or to disorient 48 times the number of bees present in the colony. Now, of course there are caveats in making such conclusions, and in generalizing to citrus (annual v perennial, differences in flowering behavior and treatment timeframe, acute exposure v low-dose chronic exposure, etc...), but there is, to my thinking, more than enough here and in other studies to suggest extreme caution is appropriate, especially when you consider that much of this pollen is destined for the queen and the larvae, who may well be even more susceptible to pesticides than adult bees.
http://pierreterre.com/page/impact-imidacloprid-bees

I think Skinns approach is sound.

Do not apply till after flowering is finished and fruit are small.

Do not apply to those that bloom year round.

Imid was singled out as "the" villian in CCD by many reports. I personally believe that the selective breeding of bees is probably a major part of the problem. There has been some research that indicates that old school bees seem to thrive better than the new and improved super bees marketed today. Also there are less mite problems with the unimproved smaller bees. This is the direction that I hope to move in when I can. Look at Top Bar Hives which promotes better bee housekeeping.

Do I want to use any insecticides? No

BUT

the reality is that in some cases you have few reasonable options.

elsedgwick,

I agree that the data you refer to lead to opposite conclusions than what I referred to. I certainly cannot judge on the validity of research A or B but I am with you on that it is better to be safe.

However, as Skinn and Darkman pointed out, applications of imid after bloom should mean no harm to bees, right? In our region, citrus blooms in March, maybe early April. Clm do not normally show up before May. So, there are no reasons to apply imid before end of April... Therefore, no harm to bees, right?

I agree that there may be something to the issue of timing, and that's what I was referring to when I mentioned the caveats in applying those studies to citrus, with the differences in flowering behavior and treatment timeframe. However, I haven't seen references to good studies on the persistence of the neonicitinoids in the plants, and particularly in nectar and pollen. Of course, there is some anecdotal evidence along the lines of "well, I tend to start seeing CLM damage x weeks after the last application", but it's still not. Basically, I think that there may be such a thing as appropriate usage of imidacloprid; but I know there has been, is, and will continue to be misusage of it.
Additionally, while CLM damage is unsightly, I'm just not convinced it effects mature trees significantly, and young ones are easily protected using other methods (including mechanical control, placement - CLM don't seem to affect plants placed along western walls to the same extent as trees in shade, e.g.- although I don't have empirical evidence to back this up, and oils). Plus, imidacloprid also kills the orange dog caterpillars; and I don't mind losing some foliage if it means more giant swallowtails around (and this year there have been tons; it's great).
Of course, HLB might shift the equation, but thankfully it is not yet in widespread evidence here in N. Florida/S. Georgia - while there have been isolated cases, they seem to have been in trees imported form further south, and I have yet to see a single ACP. Studies vary on the reported cold hardiness of ACP, but most I've seen report significant ACP mortality at 32 F/0C, so perhaps our freezes and the relatively low concentrations of citrus here will afford some protection.