Comparison of the transfer of coumaphos from beeswax into syrup and honey1

The organophosphate insecticide coumaphos has recently received emergency approval in the United States for control of fluvalinate-resistant Varroa destructor and the small hive beetle, Aethina tumida Murray. We investigated the transfer of coumaphos from wax into syrup and honey, using adsorption of coumaphos from diluted syrup or honey onto a solid-phase extraction cartridge, elution, and subsequent analysis. Coumaphos in syrup was quantitated using HPLC with UV detection, and we found that coumaphos migrates from wax into syrup, with low concentrations increasing over a few months. Concentrations reached 200–300 ppb in 100 g of syrup in contact with 10 g of wax containing 1000 ppm of coumaphos; contact with wax containing 100 and 10 ppm led to lower amounts. Impurities made HPLC determination of coumaphos in honey impossible, but the solid phase extract could be analyzed by gas chromatography/mass spectrometry. Concentrations in honey were similar to those in syrup, reaching 430 ppb after 26 weeks at 1000 ppm in wax. coumaphos / wax / honey / residues / pesticide contamination Apidologie 32 (2001) 119–125 119 © INRA/DIB-AGIB/EDP Sciences, 2001 * Correspondence and reprints E-mail: [email protected] 1 Mention of trade names or commercial products in this article is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the US Department of Agriculture. J. Kochansky et al. 120 coumaphos within the colony and the effects on bees were studied by van Buren et al. (1992a). The organophosphate insecticide coumaphos has recently received emergency approval in the United States for control of fluvalinate-resistant V. destructor and the small hive beetle, Aethina tumida Murray. Since honey is used for human consumption, several groups have conducted coumaphos residue analyses on wax and honey. Neuhauser and Krieger (1988) claimed that residues in honey did not exceed 10 μg/kg (parts per billion, ppb, parts per 109) even after several years’ use. Thrasyvoulou and Pappas (1988) found concentrations in Greek honey of up to 6 ppb and in wax of up to 2.83 mg/kg (parts per million, ppm, parts per 106). Similar residue levels were found in honey by others (Gallo and Genduso, 1986; Van Buren et al., 1992b; Fernandez Garcia et al., 1994; Garcia et al., 1996), and but were not detected in honey by Fernandez Muiño et al. (1997). In the laboratory, Wallner (1992) showed that coumaphos could be transferred into honey at detectable levels (0.5 ppb) from wax containing as little as 1 ppm of coumaphos, with concentrations increasing with higher wax concentrations. A later paper (Wallner, 1995) provided further data; concentrations of coumaphos in honey of from 0.7 to 94 ppb resulted from contact of honey with thin layers of wax containing from 1–400 ppm of coumaphos in 30 days at 30 °C. Analysis was carried out by solid-phase extraction and gas chromatography, but experimental details beyond that were not given. We decided to investigate the transfer of coumaphos from contaminated wax into syrup and honey under laboratory conditions. 2. MATERIALS AND METHODS Syrup was prepared from SigmaUltra grade sucrose (Sigma). A syrup concentration of 67% w/w (sucrose:water 2:1 by weight, ‘2:1 syrup’) was used. Unless otherwise specified, references to ‘syrup’ refer to this formulation. Honey was a blend of several commercial products obtained from a local grocery store. Analysis of this blend prior to experiments showed no coumaphos. Coumaphos was an analytical standard from Chemagro (now Bayer Animal Health, Shawnee Mission, KS), nominally 98.7% pure. Methanol and acetonitrile (both HPLC grade) and oxalic acid dihydrate (ACS reagent grade) were obtained from Aldrich. Beeswax was from a supply obtained from Betterbee (Greenwich, NY) before the registration of coumaphos in the US. No evidence was found for the presence of any coumaphos in this wax (absence of any coumaphos from samples of syrup or honey in contact with control wax). Water was purified with a Nanopure apparatus (Barnstead/Thermolyne, Dubuque, IA). An initial concentrate of coumaphos was prepared by dissolving coumaphos (50 mg) in beeswax (50 g) to yield a concentration of 1 mg/g of wax (1000 ppm). This was used for the highest concentration, and dilutions of the concentrate were used for the 100 and 10 ppm concentrations, with enough untreated wax to make a total of 10.0 g in each case. The wax samples (10 g) were placed in the bottom of 250 mL Erlenmeyer flasks, melted on a hot plate, swirled to mix, and allowed to cool. Concentrations of coumaphos in the wax were not assayed directly. A model system was used instead. A solution of 1000 ppm of CI disperse blue 14 (1,4-bis(methylamino)anthraquinone) in wax was prepared; dilution of 0.5 g of this with 10 g wax gave a homogeneous light green color in a shorter time of swirling than that used for the coumaphos. Since the coumaphos was dissolved in the molten wax, it would not be expected that difficulties would occur in the mixing. In another study, coumaphos concentrations observed in wax were reported to be similar to those added (Fries et al., 1998). Transfer of coumaphos from beeswax extract solutions onto a 3 mm × 250 mm column packed with C18 silica gel (Supelcosil LC-18-DB, 5 μm particle size, Supelco, Inc., Bellefonte, PA), using a full 20 μL injector loop. For the syrup runs, flow rate was 0.7 mL/min of methanol–acetonitrile–0.01 M oxalic acid (20:30:50 by volume) (Oka et al., 1994). The oxalic acid was necessary for other analyses being carried out at the same time, and caused no problems for the coumaphos analyses reported here. Coumaphos was detected at a wavelength of 315 nm with a SpectraSYSTEM UV2000 detector and quantitated with an SP4400 integrator (Thermo Separation Products, San Jose, CA). Range was 1.0 absorbance units full scale for all analyses. Each point is the average of three injections. Injections of a series of coumaphos standards in the range of 0.2 ng–2 μg gave a linear plot with an average detector response of 6438 detector units/ng. This technique could not be used for analyses in honey, since some of the honey pigments coeluted with the coumaphos and made quantitation difficult. Coumaphos in honey was therefore analyzed by gas chromatography/mass spectrometry using a similar solid-phase extraction, but the final elution was with 1 mL of toluene. An aliquot of the toluene fraction (1.0 μL) was injected by an on-column injector (J&W Scientific, Folsom, CA) onto a Restek Rtx-5 MS capillary column (30 meter, 0.25 mm ID, 0.25 μm df, column head pressure 0.9 bar, Restek Corporation, Bellefonte, PA). The initial column temperature of 60 °C was held for 1 minute, then the temperature was raised 20 °C/min to 270 °C and held at 270 °C for 14.5 minutes (total run time 26 minutes). The column effluent was analyzed on a Finnegan GCQ mass spectrometer (Finnegan MAT, San Jose, CA) operated in positive electron impact mode (70 eV) at a source temperature of 165 °C. The transfer line between the GC and the source was held at 275 °C. The amount of coumaphos was determined by comparison of sums of ions at m/z 362, 334, and 306 in the Syrup or honey (100 g) was added to the flasks, which were stored in an incubator at 34 °C in the dark. Flasks were not agitated during incubation. In the case of the honey samples, two flasks of each concentration flasks were prepared, and were sampled alternately at successive sampling times. Two sets of flasks were used to extend the duration of the experiment without changing the honey/wax ratio any more than necessary. In the syrup study, only 3 samples were taken over the 20 weeks, whereas in the honey study, 11 samples were taken over 26 weeks. Available incubator space limited the number of flasks we could use. Syrup samples were taken over 20 weeks before the samples developed mold; honey samples were taken over 26 weeks before the viscosity of the honey became too high for convenient sampling. Flasks were sampled using a pipette to withdraw aliquots and weighing out 5.0 g of sample from the pipette into a 50-mL Erlenmeyer flask. Coumaphos was isolated from the samples using solid-phase extraction cartridges (Oasis HLB, 200 mg/6 mL, Waters, Inc., Medford, MA), with the aid of a vacuum manifold (Pierce, Rockford, IL). Cartridges were conditioned with methanol (1 mL), followed by water (1 mL). Honey or syrup (5 g) was diluted with water (10 mL) and passed through the conditioned cartridge. After passage of 5% methanol (1 mL), followed by 100% methanol (1 mL), the coumaphos was eluted with acetonitrile (1mL). The 100% methanol and acetonitrile fractions were collected separately, as the methanol fraction sometimes contained a little coumaphos. The two fractions were analyzed separately, and coumaphos concentrations were added to obtain the final result. We did not investigate adsorption of coumaphos onto the walls of the various glass and plastic objects involved in the assay. Coumaphos was determined in syrup extracts by high performance liquid chromatography (HPLC) by direct injection of 121