Honey Bees: Estimating the Environmental Impact of Chemicals - Chapter 12 pptx

12 Typology of French acaciahoneys based on their concentrations in metallic and nonmetallic elements J.. Subirana Summary The elemental analysis of 150 French acacia honeys Robinia pseu

12 Typology of French acacia

honeys based on their

concentrations in metallic

and nonmetallic elements

J Devillers, J.C Doré, C Viel,

M Marenco, F Poirier-Duchêne,

N Galand, and M Subirana

Summary

The elemental analysis of 150 French acacia honeys (Robinia pseudoacacia

L.) collected by beekeepers in apparently polluted and nonpolluted ronments was performed by using inductively coupled plasma atomicemission spectrometry (ICP-AES) to measure significant concentrations

envi-of Ag, Ca, Cr, Co, Cu, Fe, Li, Mg, Mn, Mo, P, S, Zn, Al, Cd, Hg, Ni, and

Pb Fortunately, Cd, Hg, Ni, and Pb were not detected in the analyzedsamples Conversely, Ag, Cu, Al, Zn, and S were found in some sampleslocated near industrial areas Because a high variability was found in theconcentration profiles, correspondence factor analysis was used to ratio-nalize the data and provide a typology of the honeys based on the concen-tration of these different elements in the honeys The results wereconfirmed by means of principal component analysis and hierarchicalcluster analysis Finally, the usefulness of the acacia honey as a bioindica-tor of heavy metal contamination is discussed

Introduction

The continued expansion of industrial production and the growing use ofchemicals in agriculture have led to an increase in the number and quanti-ties of xenobiotics released into the different compartments of the bios-phere [1] The health risks to human and nonhuman biota associated withthese chemicals are evaluated on the basis of critical and reliable informa-tion on exposures and on related adverse health effects [2] In this process,the estimation of the environmental concentrations of the hazardouschemicals plays a key role A number of precise technical samplingmethods are available for monitoring pollutants in the environment.However, due to their high technicality and cost, they are generally notused routinely [2] Conversely, bioindicators are now widely employed forestimating, at low cost, the level of contamination of organic and inorganicchemicals in aquatic and terrestrial ecosystems [e.g 3–5]

Thus, honey bees commonly forage within 1.5 km of their hive andexceptionally as far as 10 to 12 km, depending on their need for food andits availability [6] During their foraging flights, they visit numerous plants

to gather nectar, pollen, honeydew, sap, and water Honey bees also visitpuddles, ponds, and other aquatic resources to collect the 10 to 40 liters ofwater which are necessary annually for the colony [7] When honey beessettle on leaves, penetrate in the corolla of flowers to gather nutritive sub-stances, and collect water in aquatic resources, they provide compositesamples from thousands of different visited points spread across a broadarea Consequently, these insects and their products such as honey, wax,

or royal jelly can provide a good idea of the level of contamination whichcan be found in air, soil, vegetation, and water in a radius of a few kilome-ters from their hive [8, 9]

Heavy metals, which are ubiquitous environmental pollutants, arefound in all the compartments of the biosphere and in living species [e.g.10–13], including honey bees and their products [14–24] In this context,

samples of French acacia (Robinia pseudoacacia L.) honeys, directly

col-lected by beekeepers in hives located in media presenting differentdegrees of pollution, were analyzed for their concentrations of heavymetals and some other metallic and nonmetallic elements in order to seewhether it was possible to find a relationship between industrialization andthe levels of honey contamination by heavy metals and related com-pounds An attempt was also made to provide a typology of the honeysamples from the multivariate analysis of their concentrations of metallicand nonmetallic elements in relation to environmental variables

Materials and methods

Sampling

Under the authority of the CNDA (National Center for the Development

of Apiculture), beekeepers of various French departments were first tacted by letter to determine their interest in being involved in a studydealing with the elemental analysis of acacia honeys and their typology onthe basis of environmental variables A sampling protocol and material tocollect and store the honey were then sent only to those beekeepers inter-ested in the project and who agreed to provide all the necessary informa-tion to interpret the analytical results found with their honey(s) In theprotocol, beekeepers were required to select one hive located in an unpol-luted area and another near a source of pollution such as an industry,mine, highway, urban area, and so on It was necessary to manually collectthe honey samples by slow extraction from the combs Beekeepers had touse the material provided for the study to avoid problems of external cont-amination by trace elements The use of bee smokers was prohibited, and

con-it was also forbidden to smoke during the sampling process Honey

Elemental analysis of French acacia honey 249

samples had to be stored in small hermetically sealed containers whichwere certified as free of trace elements, and were sent out to the bee-keepers.

The environmental conditions around the hives had to be clearlydescribed It was also required to give some climatic information, such asthe main direction of the winds, and so on If the two hives selected by abeekeeper were located in the same department, the kilometric distancebetween them had to be provided Finally, any unusual event (e.g fire)also had to be mentioned

A total of 150 different acacia honeys were obtained from variousFrench departments (Figure 12.1) All samples were collected inMay–June 1999 Honeys were sent by post to the analytical laboratory fordetermination of their metallic and nonmetallic element content

250 J Devillers et al.

Figure 12.1 Honey sampling regions in France (in dark).

Analytical method

Prior to the preparation and chemical analysis of the honeys, the sampleswere coded and randomized to avoid identification of their location andcharacteristics by the chemists The mineralization of the honey sampleswas performed in polypropylene-stoppered vials of volume 10 ml [Plas-tiques Gosselin, ref TR 95 PPN 10TT (vials) and ref B135 (stoppers)] by

ref 408071) The nitric acid was diluted in a 2/3 ratio with water previouslypurified according to the guidelines of the French Pharmacopoeia (10thedition) For each honey sample, amounts of 1 g and 2 g, exactly weighed,were digested with 5 ml of the above acidic solution Stoppered vials wereplaced in a bain-marie and warmed up to the temperature of mineraliza-tion of 60°C After 3 to 4 hours under these experimental conditions, thevolume of each vial was exactly adjusted to 10 ml with HNO3(2/3) and themineralization at 60°C was continued as described above The timerequired to obtain complete mineralization of a sample ranged from 6 to 7hours and the product was analyzed after keeping it for 15 hours at roomtemperature A solution of 5 ml was injected into an inductively coupledplasma atomic emission spectrometer (Panorama, Jobin & Yvon) previ-ously calibrated for the 18 metallic and nonmetallic elements studied Thezero point was obtained from the acidic solution used to mineralize thehoney and which corresponded with a blank The wavelengths (nm) of theemission peaks of the 18 elements studied were the following: aluminum(Al), 396.152; cadmium (Cd), 226.502; calcium (Ca), 317.933; chromium(Cr), 267.716; cobalt (Co), 228.616; copper (Cu), 324.754; iron (Fe),259.940; lead (Pb), 220.353; lithium (Li), 670.776; magnesium (Mg),279.553; manganese (Mn), 257.610; mercury (Hg), 184.887; molybdenum(Mo), 202.032; nickel (Ni), 231.604; phosphorus (P), 178.225; silver (Ag),328.068; sulfur (S), 180.672; zinc (Zn), 213.856 All samples were analyzedautomatically in triplicate by using the spectrometer In addition, for eachsample, both quantities (i.e 1 and 2 g) were analyzed The standard devia-tions were always less than 5 percent The limit of the detection of S, Al,

Ni, Ca, Mg, P, and Pb in the honey samples was 1 ng/g That for Hg was0.5 ng/g while Ag, Cr, Fe, Li, and Mn were not detected at a concentrationless than 0.2 ng/g The limit of detection of Co, Cu, Mo, Cd, and Zn was0.1 ng/g

Data analysis

Statistical analyses were performed with ADE-4 [25], a powerful statisticalsoftware program designed specifically for the analysis of environmentaldata ADE-4 includes the main linear multivariate analyses and numerousgraphical tools for optimal data display

Elemental analysis of French acacia honey 251

Analytical results

The elemental analyses obtained from 1 or 2 g of honey yielded similarresults, and hence were averaged The number of positive responses (i.e.concentrations greater than the different limits of detection) for eachmetallic or nonmetallic element in the 150 honeys analyzed and their cor-responding average, smallest, and highest concentrations (in mg/kg to raw(wet) weight) are given in Table 12.1 Detailed analytical results are listed

in Table 12.2, except for elements with a frequency of positive responsesless than 5 percent

Table 12.1 shows that calcium (Ca), magnesium (Mg), and phosphorus(P) were detected in all the samples analyzed The concentrations of thesethree elements show Gaussian distributions (graphs not given) The resultsobtained are not surprising because of the nature, role, and ubiquity ofthese fundamental elements Manganese (Mn), is also significantly present

in most of the honey samples Aluminum (Al), molybdenum (Mo), andsulfur (S) have been detected in more than 50 percent of the samples, and

to a lesser extent, copper (Cu) and zinc (Zn) About 30 percent of the lyzed samples include measurable concentrations of cobalt (Co) whileabout 20 percent of the honeys are contaminated with quantifiable concen-trations of chromium (Cr) Table 12.1 shows that silver (Ag) has beendetected in 10 samples with concentrations ranging from 0.08 to 2.16 ppm.Lithium was only measured in samples 6, 43, 44, 133, and 149 (Table 12.2)

ana-252 J Devillers et al.

Table 12.1 Number of positive responses (Nb/150) for the 18 elements studied with

their corresponding mean, lowest, and highest concentrations (in ppm)

Elemental analysis of French acacia honey 253

Table 12.2 Element concentrations (ppm) in acacia honeys collected in France

Elemental analysis of French acacia honey 255

with concentrations of 0.06, 0.06, 0.04, 0.24, and 0.02 mg/kg, respectively.Finally, nickel (Ni), mercury (Hg), cadmium (Cd), and lead (Pb), whichare particularly hazardous for biota and are indicators of industrial pollu-

surprising because about 50 percent of the samples were collected in hiveslocated in polluted areas Even if we can assume that some hives were mis-classified by the beekeepers, the descriptions provided for most of themclearly show that numerous hives were undoubtedly located near sources

of industrial pollution (e.g highways, petroleum industries) In addition,the detectable presence of some elements such as Ag or Cr clearly revealsthat some honey samples were collected in polluted areas Information onthe level of contamination of French honeys by heavy metals and relatedpollutants is scarce Recently, Fléché and co-workers [7] revealed that,between 1986 and 1996, among the routine analyses performed by theCNEVA (Centre National d’Etudes Vétérinaires et Alimentaires –National Center for Veterinary and Alimentary Studies) on honeys ofvarious origins, only 97 were focused on the detection of heavy metals,while 615 analyses were carried out for detecting pesticides and 341 wereperformed to find the level of contamination of honeys in antibiotics Inaddition, among these 97 analyses, while the presence of Pb was investi-gated systematically (with 10.3 percent positive response (p.r.)) and that of

Cd was searched in 83 samples (1.2 percent p.r.), the contamination in Hg

was only investigated in four honey samples (0 percent p.r.) Fléché et al.

[7] also emphasized that in the framework of their annual control of thequality of honeys, in 1994, the CNEVA analyzed 122 French honeys and

28 foreign honeys for their concentrations of Pb and Cd While Pb was notdetected in the former group, 43 percent of the latter were contaminated

by detectable concentrations of this element with a mean concentration of3.8 ppm Conversely, Cd was not detected in the foreign honeys while 3percent of the French honeys were contaminated by detectable amounts of

Cd with a mean concentration of 0.07 ppm [7] However, in these cal results, the type of honey was not given even though it is well knownthat this parameter widely influences the levels of contamination found insamples gathered in the same geographical area Thus, for example, in arecent study, Barisic and co-workers [24] showed that the concentrations

analyti-of Pb in meadow honey, mixed meadow and honeydew honey, and

typology of the acacia honeys based on their detectable concentrations inmetallic and nonmetallic elements, different linear multivariate analyses

256 J Devillers et al.

Multivariate analysis of the honey samples

Correspondence factor analysis

Background

Among the different linear multivariate methods that can be used to

representation of the variables and objects which greatly facilitates theinterpretation of the graphical displays [26] In addition, CFA has beenused successfully on similar data matrices for rationalizing (eco)toxicologi-cal information [27–30]

Analysis of the factorial map F 1 F 2

CFA allows the dimensionality of the 13150 data matrix (Table 12.2) to

be significantly reduced since the six first axes (i.e F1to F6) account forabout 93 percent of the total inertia of the system

The factorial map F1F2(Figure 12.2), which accounts for most of thevariance of the system (i.e 62.23 percent), clearly reveals an oppositionbetween the presence or the absence of detectable concentrations of sulfur(S) in the samples Thus, broadly speaking, the honey samples belonging

to the compact cluster of points located on the right of Figure 12.2B do nothave sulfur Conversely, points located in the top left of Figure 12.2B dealwith honey samples containing significant concentrations of sulfur It isclear that CFA can be used to perform a more precise analysis of thepoints displayed on the factorial map Thus, for example, sample number

41 does not contain a detectable concentration of sulfur but, in addition, itpresents the highest concentration in zinc (i.e 5.96 ppm) This explains itslocation as an outlier in the lower part of Figure 12.2B Conversely,samples 85 and 86 contain fairly similar concentrations of sulfur but theformer is also contaminated by Cr, Cu, and Al while the latter does nothave detectable concentrations of these elements In addition, sample 86contains more Ca than sample number 85 These chemical differencesexplain their different locations on Figure 12.2B

The strong opposition between the honeys with or “without” sulfurclearly reveals that this element has to be viewed as a contaminant It isdifficult to explain the origin of this contamination It is assumed thatenvironmental pollutions mainly explain the fairly high concentrationsfound in the honeys but direct human contamination cannot be excludedfor some samples Thus, for example, honey sample number 20 with35.90 mg/kg of sulfur was collected near a highway, as were samplesnumber 27 (S60.11mg/kg), number 85 (S18.66mg/kg), number 86(S26.05mg/kg), and others In the same way, the honey sample number

Elemental analysis of French acacia honey 257