This short report is a summary of RIRDC project GK1A conducted by Graham Kleinschmidt. Contact details: PO Box 109, Moffat Beach, Qld 4551, Phone: (07) 5491 8389 Fax: (07) 5491 9062

 
 This project was designed to help keep honey's image as a quality natural product by identifying production and transport procedures that might vary honey metal levels and adversely affect quality assurance certification and product image.

Galvanised closed head 200 1. drums have been used for many years and with new containers becoming available, the metal-related quality implications of current and alternative containers and apiary procedures were examined.

Research procedures

Metal drums for trial were selected following physical studies of galvanised, zincalume, stainless steel and plastic 200 1. drums. The trial containers were 200 1., 430 and 304/315 stainless steel, zincalume, galvanised waxed and galvanised drums.

The waxed drums were coated by Adelaide Drum Services using beeswax processed by Redpath Beekeeping Supplies. The beeswax was sampled and analysed by the Department of Agriculture, Victoria and the details are given in Table 6.

The waxed washed drums were exposed to the same washing procedure the apiarist or packer used for galvanised drums.

Six apiarists selected by H.C.A., four by Leabrook Farms and R & E McDonald incorporated the drum treatments in their commercial operations and collected both comb and liquid samples in acid washed sample containers during honey extraction (B.F.). The containers were provided by the analytical laboratories.

Replications of the treatments were sampled at Packing Houses by:-

• Honey Corporation of Australia, Brisbane
• Leabrook Farms, Adelaide
• R. McDonald/R. Goodman, Department of Agriculture in Victoria.
HCA provided the galvanised and waxed drums included in the treatments under their control. Each drum was sampled on arrival at the Packer (PO), stored for three months and then heated, rolled and resampled (P3). After pouring the drums were used as normal until the sample run one year after the previous sample.

Leabrook Farms apiarists and McDonald provided their own drums for the galvanised and waxed treatments. The sampling procedure was the same as outlined for HCA except that the storage period was 6û8 months (P6).

In addition to the above 11 apiarists who sampled over the three year period, 8 apiarists provided random spot and specific samples to assist interpretation of mainstream results. Samples from the extractor, first drum from the settling tank, mid-drum and last drum were taken to target zinc from galvanised tanks. Aluminium from centrifuges was studied by collecting samples from the extractor, the sump and post centrifuge. Aluminium was also studied in comb fumigation with and without trays, in aluminium painted and non aluminium hive-bodies and in drip trays.

A total of 912 samples (835 honeyù77 beeswax) were collected in acid washed sample containers provided by the analytical laboratories. Most of the samples were analysed for 11 metals, aluminium, chromium, nickel, zinc, copper, lead, cadmium, iron, mercury, arsenic and selenium. The eleven metal analyses were carried out by the Queensland Department of Primary Industries Animal Research Institute ICP-MS laboratory using the ICP-MS4 method. The seven metal analyses were by the State chemistry Laboratory, Victoria.

Data was statistically analysed by the University of Queensland, Gatton (UQG) Statistician and their interpretation by the Statistician and Director of the UQG Centre for Pesticide Application and Safety.

Key findings

Metal levels in all honey samples were below the maximum permitted concentration (MPC) with the levels of dangerous metals being very low, often below the level of detection. There was some fluctuation of zinc, iron and aluminium levels below the MPC.

Zinc levels were lowest in stainless steel drums and food lined bulk containers, slightly higher in beeswax lined galvanised drums and significantly (P<0.05) increased in unlined galvanised drums. The length of storage and honey type interaction influenced zinc levels and this reaction can make galvanised containers unsuitable for long term storage if the products natural image is to be maintained. Following early progress reports major packers have announced plans to phase out galvanised containers.

The mean increase in iron was low but occasional higher levels were recorded from metal drums. The current container replacement program will eliminate the iron variations.

Higher aluminium levels were recorded from some apiaries in 1994 and after modification of comb storage procedures mean levels decreased from 7.8 ppm in 1994 to 1.82 ppm in 1996.

Metal levels were higher in beeswax than honey but there was no evidence of translocation from wax to honey. Old comb wax and wax processed in old equipment contained significantly higher metal levels and should not be used in foundation comb for cut comb production.

Results and discussion

Drum trials were carried out over a three year period to evaluate lining performance over this time span. Lining durability over longer periods has not been evaluated and users would have to monitor performance to determine when performance became inadequate.

Liquid honey

Mean metal levels for all honey samples at the time of beekeeper fill (BF), at arrival at the packing house (PO) and after either three or six months storage at the packing location (P3 or P6) are given in Table 1. All eleven metals in the suite of analyses (aluminium, chromium, nickel, zinc, copper, lead, cadmium, iron, arsenic, mercury and selenium) were at levels well below the maximum permitted concentration in food (mg kg calculated as the metal, Standard A12, Metals and Contaminants in Food, June 1992) (MPC). There was some variation in the levels of aluminium, zinc and iron below the MPC whereas other metals remained at very low levels throughout.

Zinc

Zinc levels of 1994 collections varied between containers, the stainless steel treatments were very low (mean 1.20 ppm over a restricted treatment range) followed by waxed galvanised drums (1.07û4.75 ppm), washed galvanised drums (6.68û38.73 ppm) and unwashed galvanised drums (15.30û56.70 ppm). At P3 the waxed washed (WW) and waxed not washed (WNW) drum honey zinc concentration was not significantly higher but the galvanised washed (GW) drum honey was significantly higher (statistical means WW 0.409, WNW 0.424, GW 1.262, L.S.D. 0.92).

The honey from different beekeepers (and in two instances the source change in a replication) was significantly (P <0.05) different at PO sampling. The differences would be a combination of initial level and honey treatment interaction between BF and PO and this is illustrated in Table 1 and by floral zinc levels such as South Australia lucerne as low as 0.56 ppm at beekeeper fill, a natural blend of honey 7.10 ppm and the 1.96 ppm overall mean for that year.

The mean percentage variation for zinc from beekeeper fill to pouring was similar over a range of drum ages. Honey stored in galvanised drums for several years ranged from 68.20 to 108.00 ppm zinc.

The waxed treatments show a significant zinc increase by a factor of about 1.7, but this is no where near the 4.9 times increase in the galvanised washed treatment. When the National Health and Medical Research Council (NHMRC) general recommendations - Leaching of metals from Food Utensils are used as the guideline, only the stainless steel performed well, waxed drums were borderline and galvanised drums were unsatisfactory throughout the full range of floral honeys collected by the research during 1994. Some flora maintained low zinc levels in galvanised drums e.g. Yellow Mallee (1.33 ppm) whereas a Spotted Gum/Gum Top Box blend (29.60 ppm) did not perform well. In these circumstances whilst the zinc range of honey collected during 1994 was always below the MPC, the wide variation in galvanised treatments (0.57ù110.00 ppm) compared to waxed drums (0.63ù11.20 ppm) and stainless steel (0.07ù2.30 ppm), when considered in relation to NHMRC guidelines, does not support the use of unlined galvanised drums.

Zinc levels of the 1995 collections varied between containers, the stainless steel drums varied between 0.62 and 1.60 ppm, waxed galvanised drums 1.00û18.20 ppm, and galvanised drums 7.80û66.30 ppm. Waxed drums were superior to unwaxed drums (P<0.05) in both 1994 and 1995, but the durability of waxing is in doubt because:-

• zinc levels in BF were lower in 1995 than in 1994 but higher in waxed drums at P3 and P6;
• zinc levels were lower in 1995 galvanised drums than in 1994 (but much higher than in waxed drums);
• when NHMRC guidelines regarding leaching and past analyses are considered, a zinc level above 10 ppm would be high. In 1994 only 1 waxed drum was 10 ppm or higher whereas 6 drums exceeded this level in 1995 when BF was lower.
A small number of random zinc analyses of honey from pallecons (Table 5) and zincalume drums (Table 1) were in the 0.39û3.40 ppm range. This compares favourably to wax and stainless steel.

Honey that was processed through a galvanised settling tank, whilst increasing from 0.75 to 1.80 ppm zinc, remained at a low level and would not have been a significant zinc contributor.

Random supermarket honey zinc levels averaged 18.98 ppm. Levels ranged from 3.00û42.30 ppm with blended light colour honey the lowest followed by straight line honey and a dark sample was the highest. Straight line honey is stored over an extended period and the time factor increases zinc levels if stored in galvanised drums.

Beekeeper zinc levels in 1996 averaged 2.31 ppm, increased to 4.39 ppm by receipt at the packer in waxed drums, 11.92 ppm in galvanised drums and 1.36 ppm in stainless steel drums. After a minimum of three months storage waxed drum honey was 7.67 ppm, galvanised drums 26.40 ppm and stainless 1.23 ppm. Honey from a small number of zincalume drums was 1.90 ppm compared to 2.40 ppm in 1995 (Table 1). A selection of supermarket honey averaged 20.63 ppm zinc in 1996 compared to 18.98 ppm in 1995. Zinc levels were higher in the speciality lines and this would be due to longer term storage because of their irregular production pattern.

Zinc levels in waxed drums were much lower than in galvanised drums but 10% of waxed drums were 10 ppm or higher, a level that was also noted in 1995 whereas at waxing in 1994 this level was 2%.

Iron

Iron is not listed in the Food Standards or NHMRC guidelines but it is a factor that could influence public conception of honey as a natural product. The variation in iron content of honey from beekeeper fill to pour in area 1 of 1.78û2.96 ppm and area 2 of 4.20û7.20 ppm should be addressed by the industry in an attempt to minimise the difference. The difference in 1994 was largely from 2.6% of samples indicating that it is a minor occurrence that could be curtailed by appropriate action within the apiarists quality control program. Waxing did not influence iron levels and "the most sensible conclusion to draw is that while different batches of honey vary in their initial iron concentration, this level is not influenced by the storage treatments that were examined."

Iron levels in 1995 were similar to 1994 and whilst low, the BF and P3 differences indicate that from a natural product image viewpoint more rigorous drum inspections within the apiarist's quality assurance program would be beneficial.

Both iron and zinc concentrations vary significantly (P<0.05) at apiary level and this reflects differences between sources and initial source/container reaction. Whilst iron concentrations appear to be largely uninfluenced by storage method, there were highly significant (P<0.01) treatment affects for zinc.

Aluminium

Aluminium is not listed in the Food Standards or NHMRC guidelines but its public image indicates possible future concern that warrants industry pro-activity. Aluminium mean levels varied from 0.64 to 66.10 ppm with a range from 0.43 to 110.00 ppm. The variations occurred only in a few apiaries and were both within the same extraction and a progressive reduction from early season to late season extraction. A separate study that included both the drum research and other apiaries found that the higher levels occurred in apiaries where breakdown residues from comb fumigation may have contacted combs prior to their placement in hives. Apiarists using fumigation but who attempted to capture breakdown products averaged 6.31 ppm aluminium, non capture apiarists 14.50 ppm and non users 2.90 ppm.

This study also investigated centrifuges, drip trays and aluminium paint and analyses indicate that their contribution to the mean (if any) was minor when compared to fumigation.

Aluminium levels have progressively decreased during the research from 7.80 ppm in 1994 to 2.03 ppm in 1996/97 (B.F. Table 1). This decrease would have been influenced by apiarist awareness and consequent modification of comb storage procedures. Aluminium level variation from centrifuge plants was less than 2.00 ppm.

Several of the aluminium levels in this report could have been influenced by pre honey production activities and it could be anticipated that levels will decrease as the current industry trend to cold storage instead of fumigation gains momentum.

Honey comb

Mean 1994 comb honey and beekeeper fill metal analyses were similar, indicating that extracting processes did not have a significant influence - comb honey aluminium 6.84 ppm whereas beekeeper fill was 6.01 ppm, zinc 1.76 and 1.96 ppm and iron 3.63 and 3.52 ppm.

There were considerable differences in metal levels between the honey and wax in the same comb, the aluminium honey level mean was 6.88 ppm whereas the beeswax was 18.06 ppm, zinc 1.76 and 10.00 ppm and iron 3.12 and 9.02 ppm. Analyses do not indicate a significant leaching of metal from wax to honey in that cell.

The aluminium levels of beeswax foundation ranged from 2.20 to 14.00 ppm with a mean of 4.65 ppm, zinc 2.39û41.30 ppm with a mean of 14.58 ppm, iron 3.70û54.40 ppm with a mean of 8.70 ppm and copper 0.28û5.57 with a mean of 2.57 ppm. Culled combs processed in an old galvanised drum with copper coils contained 90.80 ppm zinc, 12.50 ppm copper and 119.00 ppm iron, indicating that wax processing containers may influence wax metal levels. The combination of a range of beeswax sources provided beeswax for foundation that in this instance contained lower metal levels than comb, and it is probable that careful wax processing procedures could further reduce these levels.

Metal levels in 1996 comb honey were similar to 1994 and 1995 with the exception of aluminium which was much lower. This could be attributed to the management changes referred to under liquid honey.

The levels of zinc, aluminium and iron in the wax portion of honey comb, whilst always under MPC, were lower in 1996 than 1994 (zinc 13.20 ppm to 4.14, aluminium 19.70 ppm to 4.25 and iron 16.60 ppm to 2.99). These reductions were probably a combination of improved seasonal conditions that promoted new wax and specific reference to wax processing at 1995 conferences.

Processed beeswax

Mercury and selenium were both <0.01 in 1994 dark coloured and 1995 light coloured processed beeswax whereas all other metals were at lower levels in the light coloured wax.

Analyses of processed beeswax and comb foundation confirm that not all beeswax is suitable for comb honey production, and that a check on the suitability of that batch of beeswax prior to hive placement could be warranted if the total final product (wax plus honey) was subject to analysis.

BF = Apiarist’s sample from extractor

P = = Sample from Packer with number indicating month’s of storage

Implications, recommendations and outcomes

1. Zinc

Whilst all honey metal levels were below their MPC, zinc level variation in some containers exceeded NHMRC recommendations. This could result in a challenge to the natural product image promoted by the industry. It might also be used to gain market advantage via technical exclusion.

Unlined galvanised drums should be progressively replaced by other suitable containers. Beeswax lined galvanised drums reduced zinc levels, but gradual increases from year 1 to year 3 indicate beeswax lining would require regular replacement. Food lined bulk containers and stainless steel drums did not increase zinc levels.

Honey that might require long term storage should not be held in galvanised containers e.g. speciality lines that are produced irregularly and have to be stored to ensure regularity of supply.

Following release of early progress reports, two large honey packers have announced plans to progressively phase out galvanised drums.

2. Aluminium

No aluminium level is set for honey but elevated levels from apiaries using comb fumigation could affect the product's natural image.

The move from comb fumigation to cold storage should be accelerated and if fumigation is necessary, should be strictly in accordance with manufacturer's directions and also ensure no breakdown residue contact with combs.

Following discussion with the apiarists involved, variation of comb storage procedures has resulted in a mean decrease from7.80 ppm to 1.82 ppm over two years and no further individual high results.

One major honey packer has notified suppliers that spot testing will be introduced.

3. Beeswax

Wax metal levels are higher than those in honey and old combs significantly higher than new wax. Dark or discoloured wax processed in old (pre stainless steel) equipment also contained high levels of some metals. There was no evidence of translocation of metal from wax to honey in the cell, therefore wax metal levels are unlikely to influence bulk honey but they could adversely affect honey comb.

Old and especially discoloured beeswax should not be recycled within the comb section of the industry and the replacement of old wax processing equipment accelerated.

The metal content of wax to be used for chunk honey or comb sections should be monitored prior to placement in the hive.

4. Industry Action

The industry has become very conscious of honey quality and as well as the change in containers, two major packers have also flagged in-house quality monitoring from hive to pack house.

Acknowledgments

The program on such a wide scale over three years in commercial enterprises was possible because of support and assistant from the staff of Honey Corporation of Australia, Leabrook Farms, R & E McDonald, Victorian Department of Agriculture and eleven co-operating apiarists. The central program was supported by random spot and specific sampling by Wescobee, Swan Settlers and an additional eight apiarists. The combined efforts of nineteen apiarists and their packing houses enabled the program to source samples from Queensland, New South Wales, Victoria, South Australia, West Australia and Tasmania.

Advice, assistance and analyses from both Mr H. Mawhinney, Queensland Department of Primary Industries Animal Research Institute and Mr B. Shelley, State Chemistry Laboratory, Victoria, provided strong support for the project.

Container manufacturers, Rheem and Van Leer provided assistance and co-operated in provision of several of the drum treatments.

The preliminary studies carried through by Mr R. McDonald and the apiarists who supported his efforts provided the background information and structure that was a major assistance to this project.

Research organisation

Coordinator : Mr G. Kleinschmidt

Sampling: Honey Corporation of Australia, Brisbane (HCA)
Leabrook Farms, Adelaide
R & E McDonald, Castlemaine
R. Goodman, Department of Agriculture, Victoria (DAV)

Analyses : Queensland Department of Primary Industries
ICP-MS Laboratory, Yeerongpilly (QDPI/ARI)
State Chemistry Laboratory, Victoria (SCL)