In 1995 we added the capability of monitoring volatile organic chemicals (VOC's). Air samples were collected on thermal desorption tubes within the unpainted hive boxes (brood nest and honey super), from the pollen trap hoppers, and from locations about the site being assessed. Control samples from Missoula were collected from the hive interiors and the laboratory in which the samples were stored and analyzed. We used 11.5cm x 6mm OD x 4mm ID Carbotrap 300 tubes (Supelco) with three sorption phases:

• 300 mg of 20/40 Carbotrap C - graphitized carbon black with 10m2/gram surface area for trapping and efficiently releasing molecules in the C9 to C30 range
• 200 mg of 20/40 Carbotrap B - graphitized carbon black with 100 m2/gram surface area for trapping and releasing molecules starting at the C4-C5 range
• 125 mg of 60/80 Carbosieve S-III spherical carbon molecular sieve with 820 m2/gram surface area for trapping small airborne molecules such as chloromethane.

Desorption tubes were connected with Tygon tubing to low flow sample pumps (SKC, Inc models 222-3 and 222-4) adjusted to 22.5 ml/min and 60 ml/min, respectively. The distal end of the sorption tube was protected from bee interferences and dirt by attaching a copper tube with a brass compression screw and vespel/graphite ferrule. Total volume pumped was obtained by multiplying factory-calibrated cycle volumes for each pump times the number of cycles registered on each pumps' digital cycle counter. Pumping periods were varied from eight hours to 24 hours.

Sample tubes were sealed in individual vials, placed in "whirl-pack" plastic bags and stored immediately placed in a 4 °C refrigerator. Samples from an east coast test sight were air expressed, with trip blanks on dry ice, to our University of Montana labs on dry ice. Once in Missoula, they were again stored in a 4 °C refrigerator.

Thermal Desorption/GC/MS Analysis

Air samples were analyzed by thermal desorption/GC/mass spectrometry. Sample tubes were placed in an 8-station thermal desorption unit (Dynatherm MTDU 910). After an initial helium (Liquid Air ultra high purity grade) purge for five minutes at 46 °C, tubes were subjected to a 10-minute desorption cycle at 300 °C. A five-minute cooling flush was used to remove residual contaminants trapped in the sorbent bed and transfer line. All phases of the desorption utilized a helium flow rate of 35 ml/min.

Thermally desorbed contaminants from the sample were captured by a 6 inch Vocarb 3000 trap from Supelco (10 cm Carbopack B graphitized carbon, 6 cm Carboxen 1000 molecular sieve and 1 cm 1001 molecular sieve) installed in our Tekmar LSC2000 Liquid Sample Concentrator. From there, the sample was introduced into the gas chromatograph by heating the Vocarb 3000 trap to 260 °C and flushing it with 40 ml/min of ultra high purity helium. The entire helium flow from the trap entered the GC for 15 seconds and was split 1:50 thereafter.

Chromatographic separations were accomplished on a Hewlett Packard GCD instrument containing a 60m x .32mm ID Restek RTX-502.2 capillary column (phenylmethyl polysiloxane, 1.8 mm coating). Helium flow was 1 ml/min and the total time of an analysis was 48 minutes (five min. initial temperature 40 °C, ramp 5 °C/min to 220 °C, seven minute hold time at 220 °C). Detection of the mass spectrum covered a mass range of 35 to 260 m/z, although some runs were extended to 435 m/z.

Preliminary Results

We were successful in identifying a number of VOC's in bee colonies. Some are attributable to the materials used in hive construction; some sre presumably released by honey bee colonies as physiological byproducts, of nectar or pollen; and some are contaminants from storage of samples in our laboratory refrigerators. In a few cases, we recorded peaks from pollen hopper atmospheres for organochlorine contaminants.

We captured mass spectra for a wide variety of terpene species: pinenes, camphene, ocimene and 3-carene. Terpene signatures were more pronounced in the hive interiors than in the hive pollen hoppers because the interior of the colonie contact exposed, unpainted softwood surfaces while the hoppers are largely surrounded by polyethylene, polyvinyl chloride or latex-painted wood surfaces.

A variety of VOC's were detected and identified that are common physiological intermediates and by-products, e.g., ethanol, acetone, hexanal, methyl benzoate. Since these compounds did not appear in ambient air, trip blanks, or method blanks, we presume they are characteristic of colony metabolism. The scientific literature is sparse with details on bee physiology, but we will expend some additional effort in substantiating assignment of these compounds to bee activity or their presence in plant nectars and pollens.

Storage of our samples in our chemistry lab refrigerator was responsible for introducing contamination from benzene and other hydrocarbons. A number of samples from other studies were stored in the same storage unit and consisted of benzene fish tissue extracts and diesel-contaminated soils. A very strong benzene peak appeared on most chemical traps that were stored in the chem lab refrigerator for any period of time.

Finally, we detected small amounts of chloroethane in the pollen hoppers. This could arise from contaminants at the test site. It would not be inconsistent to find them in the pollen hoppers because they are denser than air and might accumulate preferentially in the lower regions of the hive, where the pollen hoppers are located. They could also represent off-gasing of materials used to construct the pollen trap hoppers, especially since the slanted sides of the hopper were constructed from polyvinyl chloride plastic sheets. We are conducting further tests of hopper materials to clarify this question.