Honey bees are affected by a large number of parasites and pathogens. Mostly these have a set of well-defined symptoms that do not relate to CCD. For example, there are two major bacterial diseases that affect the brood: European Foul Brood (caused by Mellisococcus pluton [10]), and American Foul Brood (caused by Paenibacillus larvae [11]). There is also a fungal disease of the brood Ascosphaera apis [12]. These organisms have no effect on adult bees but have distinctive symptoms in larvae and pupae.

The parasitic mite Varroa destructor infests brood cells and lives phoretically on adult bees [13]. But heavy mite infections are obvious to professional beekeepers, especially by the stage where colonies are dying of the infestation. So in itself, Varroa infestation is unlikely to cause CCD.

A Tarsonemid mite Acarapis woodi can infest the trachea of adult bees [14] and is now widespread in North America. Acarapis infections were once thought to be the cause of the famous Isle of Wight disease, with symptoms like CCD. However, eminent honey bee pathologist L. Bailey is extremely sceptical that Isle of Wight disease has anything to do with an infectious agent [15]. This is not to say that the Isle of Wight disease is the same as CCD, nor does it exclude the possibility that Acarapis may contribute to CCD.

A protozoan, Nosema apis, infests the guts of adult bees, and when present in high numbers, causes dysentery and early senescence of adult workers [16]. This is also unlikely to be the direct cause of CCD, because the dysentery is obvious and because just about all honey bee colonies are chronically infected with the parasite every spring, even when there are no colony losses. In an interesting twist, however, a new Nosema species, N. cerana, has been recently identified from the Asian hive bee Apis cerana [17] and has now been found on A. mellifera in Europe [18–20]. This “new” pathogen has spread to the US and some researchers speculate that it has contributed to CCD.

More likely to play a role in CCD are a variety of viruses that affect adult bees (Table 1). Most adult honey bees carry symptomless viral infections [21,22]. However, under conditions of stress caused by poor nutrition, inclement weather, or parasitism by V. destructor [23] or N. apis [24], viral populations can increase and cause symptoms in adult bees. The paralysis viruses cause adult bees to tremble and shake, crawling away from the nest unable to fly. Paralysis can certainly reduce the life expectancy of workers dramatically [25], and cause spring dwindling. But in the 2007 outbreak of CCD, there was no evidence of trembling distressed workers. Therefore, the paralysis viruses are not strong candidates for the causative agent of CCD.

Anecdotal evidence suggests that CCD is more common in businesses in which bees are trucked large distances and rented for pollination.

Like other ranchers, many commercial honey producers are compelled by economic necessity to treat their livestock with a cocktail of drugs and pesticides to keep them healthy. Of particular relevance to CCD are the pesticides used to control the aforementioned brood parasite V. destructor, the cleptoparasitic small hive beetle, A. tumida, and the pest of stored combs, the wax moth G. mellonella. V. destructor was introduced into the US in the late 1980s [13]. It now infests virtually every colony nationwide [41] and has been responsible for the virtual elimination of feral colonies. (Feral colonies are now returning, because the Africanized bee is resistant to the mites [42, 43] and the mite may be losing virulence [41].) However, in the commercial setting, the mites must be controlled—usually chemically.

Apistan, containing the synthetic pyrethroid fluvalinate, is no longer effective for the control of Varroa due to the evolution of resistance [44,45]. It has been replaced with plastic strips containing the organophosphate coumophos [46]. However V. destructor has now developed resistance to coumophos as well [47], and coumophos is now being substituted by Amitraz, a triazapentadiene compound of unknown action. Beekeepers may be increasing dose rates or trying cocktails of chemicals. Some chemicals, particularly fluvalinate, may accumulate in comb wax [48], perhaps exposing commercial honey bees to levels of chemical residue that are inimical to worker longevity. Other beekeepers have tried more “organic” approaches, including fumigation with formic acid [49], oxalic acid, or essential oils [50,51]. Although these approaches do not place insecticides in colonies, they may also be less effective at controlling mites, and can be directly toxic to the bees.

American agricultural systems are dependent on the use of pesticides. Where insecticides are used, honey bee losses are common, and where bees are required for pollination, careful management is required to minimize bee losses.

To maintain effectiveness, new insecticides are constantly in development. Sometimes whole new classes of compounds are developed. Before release, all new compounds go through a rigorous registration process that includes assessment of risk to nontarget organisms, including honey bees. Insecticides must be applied in a manner that is nonhazardous to bees and other beneficial organisms. But as with all risk assessment, it is difficult to foresee all possible consequences of wide-spread usage of a particular compound. Perhaps some new insecticide-related phenomenon is now manifesting as CCD.

Bee poisoning is not very likely in early spring in the northern US, where CCD was most widely reported. Moreover, symptoms of acute insecticide poisoning—large numbers of dead and dying bees at the entrance to colonies—are easy to spot. Nonetheless, beekeepers and some scientists remain suspicious that not all new compounds are safe for bees. For instance, wide spread losses of colonies in France in recent years have been blamed on the nicotine-like insecticide Imidacloprid [26]. Imidacloprid acts on the nicotinic acetylcholine receptor of many invertebrates [27]. Because of its low mammalian toxicity, high effectiveness, and high mobility in plant and mammalian tissue, it is often used as systemic insecticide for the control of sap-sucking insects in crops and blood-sucking insects in companion animals [28]. Therein lies the possible problem for honey bees: when applied to plants the insecticide may end up in nectar or pollen.

There is considerable debate about the chances of this happening to a degree that bees are endangered. Some (mainly French) studies report residues of Imidacloprid in nectar and pollen at levels that are potentially dangerous to bees [26,29], while others (mainly North American) detected no residues [30]. Moreover, when Imidacloprid was fed to colonies in syrup or pollen at amounts likely to be found in the field, development and survival of colonies was equivalent in treated and control colonies [31], and contact with the pollen of treated corn plants had no affect on bee longevity [32].

Can we discount the possibility of nicotine-like insecticides as a contributor to CCD? Not completely. When individual bees are exposed to sub-lethal (some would say miniscule) doses of Imidacloprid, their performance in associative learning and memory tests is impaired [33–36]. Perhaps there is a certain level of exposure at which foragers have a higher chance of becoming disorientated and lost.

Farmers now have access to varieties of such staple crops as corn, cotton, canola, and soybeans, where the genome has been modified to express a bacterium-derived protein with strong insecticidal properties [37]. Crops have also been modified to express herbicide resistance genes, or insect protease inhibitors [37]. Genetically modified (GM) crops offer important environmental benefits in that the need for the application of pesticides on these crops is much reduced. But do the GM crops expressing insecticides in every cell pose a threat to foraging bees? To date, there is no strong evidence that GM crops cause acute toxicity to honey bees [38–40]. Furthermore, the involvement of GM crops in CCD seems less likely when we note that states like Illinois, with huge areas under GM crops, have not reported problems with CCD.

The honey price is currently depressed. Urbanization and more intensive agricultural practices are reducing honey yields nation wide. These twin factors lead many beekeepers to seek alternative income streams beyond honey production. Chief among these is the leasing of colonies for pollination, particularly almond pollination—a crop that is totally dependent on honey bee pollination. Many crops cause nutritional stress to the bees, or the transport or staging of colonies in holding yards may cause stress. When bees are moved out of these crops, they must feed on high quality pollen to restore body protein levels. This can be achieved by trucking the bees to a location with excellent floral resources or by feeding them. Presumably this is not always done. Anecdotal evidence suggests that CCD is more common in businesses in which bees are trucked large distances and rented for pollination.

Bees also need to feed on high-quality pollen in fall in order to produce long-lived bees that can survive winter [52]. In the US, goldenrod (Solidago virgaurea) is very important in this regard, and the flowering was poor in 2006 in the northeast. Perhaps this contributed to CCD in the following spring.

Remarkably, honey bees maintain the temperature of their brood nest within ± 0.5 °C of 34.5 °C, despite major fluctuations in ambient temperature [53]. If the brood is incubated a little outside this range, the resulting adults are normal physically, but show deficiencies in learning and memory [54,55]. Workers reared at suboptimal temperatures tend to get lost in the field, and can't perform communication dances effectively [54]. Although entirely a hypothesis, I suspect that if colonies were unable to maintain optimal brood nest temperatures, CCD-like symptoms would be apparent.