If you look at the model in equations (2), (3) and (4), you can see that they force fit the decrease in population of the bees on the neonicotinoid exposure. Because the decrease of the population in this model is only linked to pesticides, you need a very careful analysis of the confidence intervals and the quality of the fit or you can fall into the spurious correlation trap[1].

Also "The authors acknowledge that their study finds an association and doesn't prove a cause and effect link between the use of neonicotinoids and the decline of bee populations.". Which is good, it means they know the limits of this study.

The other case is that around 2005, we had a shift to produce more biofuels by doing more extensive single crop farming. This also had devastating effect on the biodiversity needed for the bees, insects and the birds.

The last point is, why bees are going very well in Australia even so Australia is the most extensive user of neonics?

The issues with bees is really a hard problem, most likely a combination of many factors, so I do not want people to jump on a single factor (the neonics) then claim victory and move to other things even so neonics are most likely not the main factor.

This year was bad for our bees, 8 hives, just 1kg honey per hive...

[0]: http://www.nature.com/ncomms/2016/160816/ncomms12459/pdf/nco... [1]: http://tylervigen.com/view_correlation?id=1597

I will miss the bees. I will not miss the people.

Seriously though, outside of a few desperate libertarians on the internet who needed not to believe it, I've never met a serious person who didn't see that likely connection.

1. Very low toxicity to mammals. Neonics are used in flea collars, and to kill lice in farm animal bedding materials for this reason. Neonoics are also very safe to apply by human applicators.

2. Neonics are systemic pesticides in plants. They are absorbed into the plant and provide long term protection against insects pests. Less pesticide drift, less danger to applicators, BUT it also gets into floral nectar, causing problems for bees and possibly other insect pollinators.

3. Neonics are cheap. I can buy 5lbs of Marathon 1G, a neonic pesiticde for about $80. Something more targeted, such as Rycar is more than $500 for a similar quantity, and Rycar is not systemic, it has to be applied several times.

There is a move towards more targeted pesticides such as Rycar, Floramite, Kontos, etc. but they are expensive, and in many cases more difficult to use. That's the darn problem. Neonics are terrible for bees, but for agriculture, they are really useful and there is no clear replacement.

There are three broad approaches to killing insects.

#1. Mechanically disrupt them. E.g., rip their heads off or squish them or similar. This may seem impractical at first, bringing to mind images of immigrant laborers crouching over the crops with magnifying glasses, identifying the bad insects and crushing them between thumb and forefinger.

That would indeed be impractical. The way you implement #1 is to find another insect that preys upon the pest insect, or that is a fatal parasite to the pest insect.

This can be a very safe method for dealing with the pest insect, because predator and parasite species are often very specific when it comes to their prey or hosts, often only going after a single species.

So why don't we use this method more? I'll cover that later.

#2. Disrupt their life processes chemically by using a pesticide that attacks some fundamental aspect of life.

This is risky and hard to get right, because most of the fundamental aspects of life that insects depend on are also the fundamental aspects of life that other life forms depend on. This means those pesticides are almost always harmful to far more species than just the pest you are trying to get rid of.

Insects are small, so you can sometimes work around most of the danger to other species by keeping the doses small enough so that they are huge in insects, but are small in other animals that eat the poisoned insects, or in animals and people that eat the crops that that the residue lingers on.

You can also try to design the pesticide so that it breaks down quickly. Insect lives are often on regular and predictable schedules, so for many species there will only be a short, predictable time when they are attacking crops. In that case, a pesticide only needs to stay effective for that time frame.

#3. Disrupt their life cycles chemically by using something that only affects the particular species that you wish to get rid of.

This is actually feasible! Insect behavior is essentially controlled by biological state machines, and state transitions are triggered and controlled by hormones. Suppose you've got a pest insect that hatches at a certain time of year, then spends a couple months living in the ground eating grubs, then on the first warm evening of summer emerges, finds a pond, and then swarms 2 to 3 meters above the north side of the pond, swarming until it finds a mate in the swarm, mates, and then lays eggs on your crop plants and then dies, and when the eggs hatch the larva eat the crops.

Each of those events will be controlled by a hormone. There will be a hormone that triggers the "leave the ground and fly to find a pond" behavior. Another will trigger the "find the north side of the pond" behavior. Yet another will invoke the "swarm at 2 to 3 meters" action, and the "find a mate action" after that. After the mating, another hormone will trigger the "lost in time, like tears...in...rain. Time to die" behavior in the males, and the "lay eggs and die" behavior in the females.

If you can make a synthetic version of that hormone that triggers the the start of the sequence, and expose the insects to it a few weeks before that first warm evening of summer, you can make them do everything early. If that is early enough that they have not yet become sexually matured, they will go through the motions, but nothing useful (from the insect point of view) will happen. You'll have effectively wiped out that whole generation in that area.

But what happens to other insects that get exposed to that hormone? Won't we also be triggering bees and other useful insects into doing things out of sequence? Nope! It turns out that hormones from one species generally don't affect other species, nor do they affect non-insects that might eat the insects.

So why do we do #2 instead of #3? The same reason we rarely do #1.

We rarely do #1 and #3 because to do them requires actually understanding the pest insect. If you want to bring in a predator or parasite for a pest insect, you need to know enough about the natural ecosystem of the pest to identify its predators and parasites, and understand their effects on it.

Similar for #3. Someone has to study the life cycle of the pest sufficiently to reverse engineer its behavior state machine to identify the behaviors that we'd like to fiddle with, and study the pest sufficiently to identify which hormone controls that behavior.

From what I understand, these studies aren't particularly easy. Someone may have to spend many years studying a particular insect to understand it enough to start hacking it's biological programming. The big pesticide companies aren't particularly interested, because #2 is a lot easier...that just takes formulating new chemicals that are generally hostile to life, and then figuring out what restrictions have to be placed on their use to kill insects without doing too much collateral damage.

Academic researchers don't do much for #1 or #3 either, because there just isn't the funding.

There is a good illustration of this in the book "Life on a Little Known Planet: A Biologist's View of Insects and Their World" by Howard Ensign Evans. He was one of the world's leading experts on parasitic wasps. Before reading his book, I did not even know that there were parasitic wasps, but in fact there are many species of them, most very tiny (head of a pin size).

He tells of an incident where there was an invasive pest, from Florida if I recall correctly, that was attacking California citrus crops. In Florida there was another insect that was either a predator of or a parasite of (I forget which) the pest. This was imported in an attempt to control the pest.

This attempt failed, and California's citrus crop suffered large losses. Many years later, researchers figured out why the imported predator/parasite did not work. It turned out that the predator/parasite species turned out to actually be two species. According to everything that scientists had observed and measured at the time, they appeared to be one species, but it turned out to be two closely related species. There were only a couple of observable differences, both subtle. One was something like one species mated slightly earlier than the other. That was easy to miss, because unless you've watched a lot of them mating, you won't be able to tell the difference between two populations whose mating windows overlap, and one population with a wider mating window. Unfortunately the other difference was that only one of the two was a predator/parasite to the California pest. All of the ones they collected to send to California were from the wrong population.

I believe (but don't recall for certain) that this predator/parasite species was a parasitic wasp. The reason no one had studied it enough to realize that it was two species was that in the US there were only two parasitic wasp experts, and they were busy with the thousands of other parasitic wasp species.

(There are a lot of species science has not gotten around to studying, or even cataloging. Evans mentions early in the book that every summer he'd set out an insect trap on his property in New England, and would routinely catch insects that were unknown in the scientific literature. He would even occasionally catch parasitic wasp species that he did not recognize).

Why were there only two parasitic wasp experts? Evans mentions that he had a promising graduate student who was interested in specializing in parasitic wasps, and Evans advised the student to find another specialty. Industry was not interested in hiring parasitic wasp experts, and universities entomology departments weren't growing so the only way he'd get an academic position as a parasitic wasp expert was to replace an existing retiring expert, and neither Evans nor the other US parasitic wasp expert were anywhere near retiring.

Personally, I find this ridiculous. Pest insects cause a tremendous amount of economic damage. Methods #1 and #3 are effective and environmentally safe ways to control them. I would think it would be well worth our while to fund anyone who is interested and willing to make a career out of studying the ecology of pest insects and of predator/parasite insects that might affect pest insects. Even if most do not lead to controlling pest insects, some would, and that should justify the cost.

As far as I have been able to find, no one keeps track of how many entomologists there are, but the Entomological Society of America has about 7000 members. If every member of the ESA was an expert in a dozen species, they would still not come close to covering all the species that are probably economically relevant in the United States.

In 1989-1990 US honey Bee colonies were at about 3.5M. Neonicotinoids were introduced in the 1990's and since then the US Bee colonies declined to an all time low in 2012 (below 2.5M colonies). Since 2012 there was a small but steady uptick to just above 2.5M colonies; however, that is still down 1M and nearly 33% total since the introduction of Neonicotinoids.

In fact, the biggest decline was from 1990-1996 when the number basically dropped from 3.5M to 2.5M. My understanding is Colony Collapse Disorder (CCD) wasn't even acknowledged by the Government until 2006, which is odd because the decrease from 1996-2006 may have only been 2.5M to ~2.35M. Then again in 2006 our Government was championing the strength and stability of the housing market, so maybe they are just not the most proficient at interpreting numbers and data.

As to wild/ferral honey bees, they may be endangered but I assure you they do exist in the US. In fact my parents just has a wild hive removed from their property resulting in 80lbs of honey (including the comb). You can check out Willy the Bee Man, he runs the largest wild bee removal business in South Florida and his website includes some videos of the removals.

Regardless, a better metric to consider re: managed hives might be annual colony losses, not total quantity of active hives. Of course beekeepers can make new hives. Thus far they are thankfully able to keep up with higher and higher losses each year (and honey and pollination prices are continuing to rise as more energy and money is put into raising bees instead of bee products).

This spike in losses is more recent than the varroa introduction. Graphing out one positive metric and pretending things haven't gone weird with honeybees in the past decade or so seems misguided.

Beekeepers tend to use annual colony losses for whatever reasons as a measure of overall apiary health. These are increasing (sometimes dramatically). For example, a commercial apiary may count 10% as normal and expected losses overwintering, while we are currently seeing upwards of 40-45% annual losses. The numbers can be replenished but this takes time and money and is the worrying unnatural trend that is being discussed. The absolute number of active hives at any given time is not necessarily a useful metric to measure declines.