In the present work the analysis of lipophilic matrices (bees, propolis and honeycombs) and a high water content matrix (honey) were performed. Methods for the analysis of OP pesticide residues in honey are well developed, but there are few reports of clean up methodologies for propolis and honeycomb. Propolis is a resin-like material which is used by bees for sealing the beehive. It is collected in the field from different sources and deposited in the hive after a chain of cooperative work of different specialized bees. Although both honeycomb and propolis are lipophilic, their average polarity is very different. Whereas honeycomb is a non-polar matrix with high molecular weight esters (MW > 700) as main constituents, the polarity range of propolis compounds is wider and their molecular weight much lower. Flavonoids, terpenoids and phenolic acids, which are common components in propolis, have an intermediate polarity. Nevertheless, the GPC method developed here for the determination of lipophilic pesticides proved to be suitable for the clean up step in the analysis of bees, propolis and honeycombs, even though the average molecular weight of propolis constituents and OP pesticides is similar. GPC on Biobeads gave clean extracts, with little matrix effect, due perhaps to the differences in polarity as discussed above.
4.1. Pyrethroids and Organophosphorus Pesticides
Coumaphos is an acaricide widely used against Varroa destructor
and its presence in the beehive is due to the specific treatment applied to them. On the other hand, chlorpyrifos is not employed in the management of the beehives, but surprisingly its residues were the most frequently detected in honey. Chlorpyrifos is the most employed OP in crop management in Uruguay and it clearly came from the environmental pollution outside the hive. Coumaphos, ethion and chlorpyrifos had also been found in Uruguayan commercial propolis samples [27
]. Also in this work coumaphos, chlorpyrifos and in a lower extent ethion were detected in raw propolis. On the other hand, chlorpyrifos was found to occur in many samples, in agreement to reports by Mullin et al
., in the USA [8
]. They found chlorpyrifos in 19 of 36 honey samples of active beehives and could not evidence any direct relationship to behavioural changes.
Regarding bees, other authors also found residues of OPs, including chlorpyrifos, in samples of dead bees associated to suspected cases of bee poisoning [8
] but in this work, no OP residues were found in bees.
Cypermethrin is widely used to control ants and in outdoor applications to protect crops. Its toxicity to honey bees has been known for 20 years. According to literature this insecticide is highly toxic to honey bees in laboratory tests, but field application at the recommended dosage do not put hives at risk. The LD50
for bees at 24 hours oral is 655.6 μg/kg [14
]. In this work cypermethrin was found in 19% of the analyzed samples of honeys at concentrations levels below the LD50
Cypermethrin residues were found in six of the twenty nine honey samples from active beehives. The concentration levels were below the LD50
. Sublethal concentrations of pyrethroids have also been studied and found to produce harmful effects on honeybees such as behavioural and physiological changes [17
It has been proved that the combination of pyrethroids and miticides acts synergically as toxic effects were detected at low concentrations suggesting that the saturation of the bees detoxification system was overloaded causing detectable noxious effects [32
Imidacloprid residue levels found in honeycomb samples from depopulated beehives are on the same order of magnitude of what has been reported as responsible for disorientation or other behavioural disorders in bees [34
], these results show that bees were in contact with this pesticide. This finding could suggest that imidacloprid is first ingested and then secreted with beeswax while bees build up the honeycomb.
On the other hand, its presence in propolis samples allows the assumption that bees transport imidacloprid, which is finally deposited in the beehive. For those reasons is possible to find its residues in different beehive products. Imidacloprid is used as a sunflower seed dressing agent and therefore, its most plausible origin in the cases under study is the sunflower plant. It has been reported in literature that imidacloprid contaminates all the parts of sunflowers at 1–20 μg/kg range [37
]. Particularly, the flowers are contaminated at an average level of approximately 10 μg/kg at the time of foraging [38
Moreover, average levels of few μg/kg of imidacloprid were measured in pollen and in sunflower nectar [39
]. In 2006, Chauzat et al
., reported residues of imidacloprid in nectar and pollen at levels that are potentially dangerous to bees [40
], while Schmuck et al
.detected no residues [2
]. The presence of imidacloprid from sunflowers has been controversial, but nowadays the recent improvements in the analytical methodology allow the detection of low levels of imidacloprid, levels that were unreachable at the beginning of the century.
It has been assumed that the homing ability and behaviour of forager bees may be severely affected by imidacloprid residues [13
]. Furthermore, this assumption is supported by reports of Italian and French researchers, who found that agrochemicals greatly influence bee behaviour and their feeding activities. Imidacloprid has an antifeedant/repellent effect but not a lethal or “knock-down” effect [3
] at concentrations 10 times lower than its LD50
. Sublethal effects of 24 μg/kg of imidacloprid in sucrose solution were noticed on olfactory learning performance [3
], and other neurological effects can be detected at 6 μg/kg mainly in bees’ orientation ability while visiting a feeding source [9
]. Although Nguyen et al
., suggested that imidacloprid seed-treated maize has no negative impact on honey bees [41
Usually, the relationship between the presence of imidacloprid and the toxic effect on bees has been restricted to the concentration in the flowering parts of the plant. Recently it has been proved that gutation drops from sunflower seedlings contains relative high amounts of imidacloprid and therefore other parts of the plant can be a source of this pesticide [42
]. Our results point in the same direction as propolis it is not collected from the flowers. Resin sources for propolis are the buds of trees and exudates from plants. The resins are transported in the mouth and legs of the foraging bees and passed in the hive where bees mix the resin with wax and seal the hive. Sunflower produces a complex resin of phenolics and terpenic compounds and it is possible to assume that sunflower resin, due to the abundance of the plant in a plantation, is a forage source for propolis [43
]. Therefore, it could be suggested that not only the flowers but also other parts of the plant, and even other plants, from where the bees collect propolis have to be considered and evaluated as a source of pesticide contamination. The presence of imidacloprid in propolis suggests both a systemic and a contact mode of action of the pesticide. Two different processes can be noticed: first, imidacloprid’s presence in the beehive means that worker bees are able to adapt to limited levels of imidacloprid in their bodies, which is excreted with the wax afterwards. Second, the presence of imidacloprid in propolis is a signal that the pesticide is included in the hive after a chain of events where not only a systemic route can be inferred but also a contact one is working, probably, leading to bee intoxication. It should be pointed out that imidacloprid was the only pesticide detected in these abandoned beehives.
Summarizing, in samples from depopulated beehives, imidacloprid and fipronil were detected. On the other hand, cypermethrin, endosulfan, chlorpyrifos and coumaphos were detected in honeys from productive beehives. From these results, the reasons for beehive depopulation in Uruguay cannot be attributed only to the presence of sublethal doses of agrochemicals in bees or their beehives, as not all the samples from the abandoned honeycombs contained pesticides. Notwithstanding, due to the overwhelming evidence of the effects of very low concentrations of imidacloprid on bees behaviour [34
], the imidacloprid levels found in these samples could be related to the bee disappearance phenomenon from beehives where this pesticide was found, as no other insecticide was detected in these samples. It has been reported that imidacloprid turns worker bees more aggressive [12
]. The presence of imidacloprid residues in honeycombs and propolis demonstrated that they are also in contact with this insecticide as they lay the agrochemical in the whole honeycomb, before abandoning it. Such behaviour could be the visible sign of the studied insecticide action on these bees [45
In this context, the role of agrochemicals on beehive depopulation can be understood as a consequence of neurological disorders that affect forager bee orientation, caused by their repeated exposure to sublethal doses of insecticides [12