HAL Id: hal-00890913 https://hal.archives-ouvertes.fr/hal-00890913 Submitted on 1 Jan 1991 HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés. Population dynamics of honey bee nucleus colonies exposed to industrial pollutants Jj Bromenshenk, Jl Gudatis, Sr Carlson, Jm Thomas, Ma Simmons To cite this version: Jj Bromenshenk, Jl Gudatis, Sr Carlson, Jm Thomas, Ma Simmons. Population dynamics of honey bee nucleus colonies exposed to industrial pollutants . Apidologie, Springer Verlag, 1991, 22 (4), pp.359-369. <hal-00890913>
Transcript
HAL Id: hal-00890913https://hal.archives-ouvertes.fr/hal-00890913
Submitted on 1 Jan 1991
HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.
Population dynamics of honey bee nucleus coloniesexposed to industrial pollutants
Jj Bromenshenk, Jl Gudatis, Sr Carlson, Jm Thomas, Ma Simmons
To cite this version:Jj Bromenshenk, Jl Gudatis, Sr Carlson, Jm Thomas, Ma Simmons. Population dynamics of honeybee nucleus colonies exposed to industrial pollutants . Apidologie, Springer Verlag, 1991, 22 (4),pp.359-369. <hal-00890913>
Population dynamics of honey bee nucleus coloniesexposed to industrial pollutants *
JJ Bromenshenk JL Gudatis SR CarlsonJM Thomas MA Simmons
1 Division of Biological Sciences, University of Montana, Missoula, Montana 59812;2 Terrestrial Sciences Section, Pacific Northwest Laboratory Richland, Washington 99352, USA
(Received 6 March 1990; accepted 25 February 1991)
Summary — Nucleus colonies (nucs) of = 4 500 honey bees (Apis mellifera L) were evaluated as analternative to full-size colonies for monitoring pollution impacts. Fifty nucs were deployed at 5 sitesalong a transect on Vashon Island, Washington. This provided a gradient of exposure to arsenic andcadmium from industrial sources. After 40 d, statistically significant differences were observedamong sites for mean mass and numbers of bees (P ≤ 0.01), honey yield (P ≤ 0.07), and arsenicand cadmium content of forager bees (P ≤ 0.001). These findings are discussed in terms of expo-sure to heavy metals and observed changes in colony dynamics, especially brood rearing andhoarding of pollen, nectar, and honey.
Apis mellifera / population dynamics / biological indicator / pollutant
INTRODUCTION
Honey bees (Apis mellifera L) are used tomonitor the distribution and impact of vari-ous hazardous chemicals, including traceelements, heavy metals, radionuclides,pesticides, and organic contaminantssuch as polychlorinated biphenyls (Fritzschand Bremer, 1975; Bromenshenk, 1979,1988; Wallwork-Barber et al, 1982; Celli,1983; Höffel and Müller, 1983; Bromen-
shenk et al, 1985; Anderson and Wojtas,1986; Morse et al, 1987).
The millions of bee colonies in the Unit-ed States offer an extensive, in-placemonitoring network. However, beekeepersare understandably reluctant to placebees in areas where they are exposed tohazardous chemicals. However, nucs, alow-cost alternative to full-size colonies,can fill such voids or provide an inexpen-sive sampling grid.
* The research described in this article was funded by the Environmental Protection Agency underassistance agreements CR-810929-01, CR-811387-01-1 and CR-810292-01-0 to the University ofMontana and a related services agreement TD 1589 with the Department of Energy under contractDE-ACO6-76RLO 1830. It has not been subjected to Agency review, and therefore does not neces-sarily reflect the views of the Agency; no official endorsement should be inferred.**
Correspondence and reprints
This research was conducted on VashonIsland near a heavily industrialized regionof Tacoma, Washington (fig 1 a). Previous-ly, we mapped exposures to arsenic andcadmium based on concentrations ofthese metals in forager bees (Bromen-shenk et al, 1985; Bromenshenk and Pres-ton, 1986). For this study, we assessedthe feasibility of using small populations ofhoney bees to identify and quantify re-
sponses associated with in situ exposuresto environmental contaminants.
MATERIALS AND METHODS
Each 4-frame nucleus hive was one-fourth thesize of a standard hive body (19.5 cm x 26.5 cmx 28 cm) and initially contained a newly matedqueen and 400 ± 17 (± 2 SD) g of bees, or =4 500 individuals. The nucs were stocked withbees from full-size hives at a Vashon Island api-ary exposed to relatively low concentrations ofcadmium and arsenic, as determined by chemi-cal residue analysis of forager bees (Bromen-shenk et al, 1985).
Bees shaken from the stock hives were ran-
domly allocated to the nucs based on a table ofrandom numbers (Sokal and Rohlf, 1981). Eachnuc was supplied with 3 375 cm2 frames con-
taining beeswax foundation and 1 375 cm2frame containing a fully drawn comb. The drawn
comb provided space for immediate storage ofpollen and nectar and for egg laying. Each unitwas immediately fed 1 500 ml of sucrose.
Fifty nucs were deployed on Vashon Islandon July 10, 1984, 3 d after their establishment.The nucs were placed at 5 sites (fig 1 a) alongthe known pollution gradient, extending north-east and downwind from a heavily industrializedregion of Commencement Bay near Tacoma,Washington.
Twelve nucs, set on stands in groups of 3,were placed at the south (#1, high exposure),mid-island (#3, medium exposure), and north(#5, low exposure) sites. Six nucs, in groups of3, were placed at the intermediate sites (#2 and#4).
Total hive mass was calculated each week.Initial and final mass were recorded for hive
components (hives and frames) and for liveadult bees, which were shaken from the framesfor weighing. A platform balance (range 0-59 kg, ± 6 g), calibrated with weights traceableto the National Bureau of Standards, determinedthe mass.
Weekly estimates of adult bee populationswere obtained from counts of frames covered bybees. Fractions of frames covered were visuallyestimated. Bees were restricted to the hives the
prior evening. Counts were conducted in thecool temperatures at dawn while the bees wereclustered. At the study’s end, both frame cover-age and mass of bees were determined.
Acetate tracings of each side of each framewere made weekly to obtain estimates of total
areas per hive of undrawn wax, partially drawnwax, drawn wax, empty cells, nectar, cappedhoney, pollen, open brood (eggs and larvae),capped brood (pupae), and numbers of droneand queen cells.
To make the tracings, one observer placedthe queen from each hive in a screen cage,then shook the bees from the frames. A secondobserver made the tracings from the frames.This procedure yielded 8 tracing records (identi-fied by frame) of 11 parameters per hive, for 50hives, at 5 sites, over 5 weekly intervals.
Accuracy and precision of tracings were
checked in the field for randomly chosen hives.To assess accuracy, the person tracing wouldmake a second, slower, more "precise" drawingof the same frame. To assess precision, a dupli-cate, rapid tracing would be immediately madeof the same frame.
Areas for each hive parameter were digitizedin the laboratory using a Ladd Research Indus-tries, Inc Graphical Digitizer® (Ladd ResearchIndustries, Burlington, VT) coupled to a Mon-roe® (Monroe Systems for Business, MorrisPlains, NJ) 1830 Programmable Printing Calcu-lator. Accuracy and precision of digitizing weredetermined for each observer and for each digi-tizing session. At the beginning of each sessionand for each set of 20 acetates, test patternsconsisting of a square and of small, medium,and large circles were digitized. Results werethen compared to control charts. Any results ex-ceeding control limits required corrective action:re-calibration and re-digitizing of any acetatesdigitized since the last "in compliance" measure-ments.
Estimates of exposure to arsenic and cadmi-um were obtained from residues in foragerbees. From screens blocking hive entrances, =50 returning forager bees per hive were aspirat-ed into polyethylene sample bags and immedi-ately frozen. Sample bee tissues were dried,ground, and digested in nitric acid. Heavy metalanalyses employed atomic absorption spectro-photometry supplemented with hydride vaporgeneration for arsenic determinations (Bromen-shenk et al, 1985; see procedures for details).
regression, and multiple range tests follow Steeland Torrie (1980) and Sokal and Rohlf (1981).
Statistical comparisons emphasize the finalobservation period because: 1), all coloniesstarted with nearly the same number of beesand the same equipment, so there were no sta-tistical differences at the start of the experiment;2), significant differences by date had to occursince some of the measured components, notpresent at the beginning of the experiment, onlyappeared 2 wk later (eg, capped brood was pro-duced, while the amount of uncapped brood cor-respondingly decreased); 3), more informationwas available for the initial and last observation
periods (eg, total hive weight, weights of beesversus hive components, and chemical residuevalues); 4), site-specific differences becamemore pronounced with time (figs, 1, 2).
RESULTS
Colony responses
The nucs were self-sustaining, despite fre-quent and invasive handling. At Site 3, aqueen flew out of an opened hive duringthe fourth week. The colony accepted a re-placement, with little apparent disruption ofactivities.
After the experiment’s end on August19, the mini-colonies were left at the sitesfor another month. Eighty d after the initialset-up, the nucs were again sampled forchemical residues, then disassembled andexamined. Forty-nine of the original 50 col-onies were still alive and appeared to be ingood condition.
Nucs facilitated quality assurance andcontrol of data. Observers could trace theframes in a nucleus hive in 10-20 min, atask that could have taken more than 1 hfor a standard hive.
Precision estimates based on 40 pairsof duplicate tracings revealed an overallmean difference of 6.1 cm2 (SD = 8.90,n = 40) between tracings. These errors
were usually small compared to the maxi-mum possible area (= 3 000 cm2) for eachmeasured aspect. For example, the small-est area was ≈ 60.77 cm2 (10% error) forthe total mean area of pollen in hives at asite. For most of the components meas-ured, mean areas per hive usually rangedfrom 300-900 cm2 (< 2% error).
Accuracy of tracing checks demonstrat-ed a mean difference between rapid andmore careful tracings of 11.3 cm2, (SD =18.95, n = 44) for all parameters. Based ona sign test, we found no consistent or sta-tistically significant positive or negativebias. The least accurate and precise esti-mates were obtained when only a few cells
existed on a frame, such as sometimes oc-curred with pollen. Errors were higher forestimates of empty cell areas (which mightactually contain eggs) than for areas of
capped brood. Scoring cells partially filledwith nectar or pollen also contributed to er-ror. However, compared to the overall are-as measured for each hive, these errorswere relatively small.
Laboratory digitizing contributed rela-
tively few errors. Accuracy, determined bymeasurements of a 375-cm2 test squareby several observers, demonstrated a neg-ative bias of 0.3%. However, this value
probably equals or exceeds the tolerancefor the referenced test square, which was
drawn on acetate using a metric straight-edge. Based on > 200 measurements ofseveral test shapes, precision coefficientsof variation (CV) ranged from 0.7-2.0%.The slightly higher CVs were obtainedfrom the smallest digitized areas, ie, those< 18 cm2.
Mass of bees and of hive parts was de-termined at the beginning and end of the40-d experiment. Although each unit be-
gan with approximately the same mass ofbees, some drifting among colonies oc-
curred. Weekly frame coverage assess-ments provided an estimate of populationsize. Measured mass of adult bees was re-
gressed against frames covered by bees(fig 1b). Frame coverage provided a rea-sonable (r = 0.962, n = 48) estimate of beemass, which we assume also indicates
population size (numbers of bees).A 1-way analysis of variance of the final
mass of bees (arcsine transformed data)at each site yielded an F ratio of 3.87 (4.43df), significant at P ≤ 0.009. The arcsinetransformation normalized the data and
yielded homogeneous variances.
Using Tukey’s multiple range test for
means, bee mass at Site 1 was statistical-
ly different at the P ≤ 0.05 level from Sites4 and 5. Bee mass at Sites 2 and 3 wasnot statistically different from Sites 1, 4,or 5.
Food stores as indicated by areas ofpollen (fig 2a), nectar (fig 2b) and honey(fig 2c) varied considerably by date. For in-dividual observation periods, differencesbetween sites in the amount of stored pol-len, nectar (uncapped cells), and total hon-ey (nectar + capped cells) were not signifi-cant (P ≤ 0.05). All colonies had stores ofpollen and honey or nectar at all observa-tion periods.
Capped honey stores tended to be low-est at Site 1 (fig 2c). A 1-way analysis ofvariance of the final amount (area) of
capped honey by site (logarithmic data
transformation) produced an F ratio of 2.39(4,43 df), "significant" at the 7% probabilitylevel. The logarithmic transformation nor-malized the data, since standard devia-tions tended to be proportional to themeans.
Capped honey was reduced by 67%percent at Site 1 compared to Sites 4 and5 (fig 2c). This change was proportionallygreater than the 40% reduction in the pop-ulation size of bees (frame coverage andmass) observed at Site 1 (fig 1 c). Yet dif-ferences in capped honey were only"significant" at the 7% probability level;whereas differences in bee mass were sig-nificant at the 1% probability level. Themost variable hive component measuredwas within-site values for capped honey;CVs averaged ≈ 69%, but ranged to
> 120% (fig 2c). Total nectar and honeywas less variable (CVs = approximately27%) than capped honey (fig 2d).
Colonies at Site 1 had fewer adult bees
(fig 1 c) and less brood (figs 1f-h) than theother sites. Colonies at Site 3 usually pro-duced the most brood (fig 1f-h), yet hadthe second-lowest final mass of adult bees
(fig 1c). Initially, brood rearing at Site 4was delayed. The bees were slow to ac-cept one of the 6 queens, and we replacedanother queen that the bees rejected.However, the final mass of adult bees atSite 4 was nearly equivalent to Site 5,which had the largest bee populations (fig1c).Wax production (based on comb build-
ing) first increased rapidly, then levelled off(fig 1e). Colonies at Site 1 lagged behindin wax production. Fourth-week declines inwax at Sites 3, 4 and 5 were unexpected.
Other hive components such as theamount of undrawn and partially drawnwax, number (area) of empty cells, andnumber of drone and queen cells providedmeasurements that were highly variableand difficult or impossible to interpret.
Exposure to heavy metals
Arsenic and cadmium concentrations in
forager bees revealed a consistent, site-
specific exposure pattern. Shortly after de-ployment, the sites consistently ranked 1 >
3 > 2 > 4 ≥ 5 from high to low arsenic con-tent of forager bees. Cadmium concentra-tions ranked 1 > 3 > 4 > 2 ≥ 5. This patternstill remained 40 d after the experimentended.
On the 40th day, a 1-way analysis ofvariance (ANOVA) for heavy metal con-centrations (untransformed data) in foragerbees at all sites indicated highly significant(P ≤ 0.001) differences among locations formean arsenic and mean cadmium.
A linear curve model, r = -0.71, n = 16 (fig 1d), represented the best fit for massof bees versus arsenic content of bees. N
equalled 16, not 48, because bee sampleswere pooled from sets of 3 hives at thetime of sampling. This limited the numberof foragers removed from a hive to 50
bees, minimizing sampling impact.There was a significant inverse relation-
ship between population size (indicated bybee mass) and arsenic exposure (indicat-ed by arsenic levels in forager bees).Based on the coefficient of determination,different arsenic exposure levels could ac-count for nearly 49% of the variation in beepopulation size.
The above conclusion was confirmed byKendall’s coefficient of rank correlation,which tests whether 2 rankings are sub-stantilly in agreement with one another
(Sokal and Rholf, 1981). The coefficient ofrank correlation was 0.483, significant atthe 1% level.
Regression analysis of cadmium versusbee mass did not yield statistically signifi-cant values. Cadmium levels, like arsenic,were highest at Sites 1 and 3. Unlike ar-
senic, cadmium concentrations in bee tis-
sues were higher at Site 4 than at Sites 2and 5.
Bees at Site 1 sustained the highest ex-posures to arsenic and cadmium. Bodyburdens of arsenic ranged from 10.1-18.5ppm (dry weight), and cadmium from 2.8-5.5 ppm. Bees at Sites 2 and 3 had inter-mediate arsenic exposures of 5.0-9.4 and6.1-10.4 ppm, respectively. Cadmium at
Site 2 varied from 1.9-2.6 ppm, and Site 3
ranged from 2.3-3.8 ppm. Lowest metalconcentrations were 3.3-5.6 ppm for ar-
senic at Sites 4 and 5, and 1.4—2.1 ppmcadmium at Site 5. Cadmium varied from1.85-3.10 ppm at Site 4. At all sites, arsen-ic and cadmium were elevated above
background levels of 0.1-1.0 ppm arsenicand < 0.04 ppm cadmium for colonies notin the exposure area (Bromenshenk et al,1985).
At Site 1, exposed to the highest con-centrations of arsenic and cadmium, adultbee populations increased slightly, thendeclined throughout the experiment. Popu-lations at Site 3, which received the sec-ond highest exposure to arsenic and cad-mium, displayed population increase anddecline (fig 1 c). Adult bee populations atthe sites of lowest exposure to arsenic andcadmium (2, 4 and 5) steadily increasedfollowing the emergence of the first brood(fig 1c).
From the emergence of first-broodadults until the experiment’s end, popula-tion growth curves at the 3 lowest expo-sure sites (2, 4, 5) were described by a lin-ear model (fig 1 c). Growth of bee
populations at the high exposure sites (1and 3) could not be described by a linearmodel (fig 1 c). A quadratic model wasused to describe Site 1 data. Site 3 couldnot be described by the same model asSite 1. The stepwise growth curve suggest-ed by the Site 3 data may be an artifact ofthe limited number of observation periods.We used a 4th order polynomial as a bio-
logically meaningless model for compari-son with other sites (no r values shown be-cause of zero degrees of freedom).
DISCUSSION AND CONCLUSION
The nuc’s small size and ease of handlingpermitted the use of tracings to obtain ac-curate and precise measures of comb
areas. Actual measurements of several as-
pects of full-size hives usually are impracti-cal, since they are so labor-intensive andtime consuming. Visual estimates of hivecomponents can reduce observation time(Jeffree, 1958), but observer bias and sub-jectivity are difficult to control. This reduc-es data comparability. Bromenshenk andLockwood-Ogan (1990) reviewed methodsfor measuring hive components, includinga new method for digitizing hive compo-nents in situ.
The depressed adult bee populationsobserved at Site 1 (the site of highest ex-posure to arsenic and cadmium) reflectedsmaller broods and rate of comb building.This site also had the smallest stores of
surplus (capped) honey.Site 3 demonstrated the second-highest
exposures to arsenic and cadmium and atthe end of the study had the second-lowest mass of bees and honey, althoughbrood rearing was higher than at othersites. Throughout the experiment, Sites 4and 5 had more bees, and honey, and low-er arsenic exposures, than any other site.
Food availability did not appear to bethe factor limiting population size at Sites 1and 3 or at any of the other sites. All colo-nies had stores of pollen, nectar, and hon-ey at all times. It seems more likely thatsmall adult populations and fewer foragersat Sites 1 and 3 led to reduced hoarding.
Reduced brood rearing apparently con-tributed somewhat to the adult population
decline at Site 1. Increased brood rearingat Site 3 might have been a density-dependent response to fewer adult bees.Pollen from the high exposure sites (1 and3) contained high concentrations of arsenicand cadmium (Bromenshenk et al, 1985).Since pollen is fed to developing brood,the decreased brood at Site 1 may haveresulted from larval ingestion of toxic pol-len.
At Site 1, arsenic in bees exceeded le-thal levels (see Bromenshenk, 1980 for areview on toxicity). Sites 2 and 3 displayedconcentrations considered poisonous or atleast hazardous, and potentially life-
shortening. Although elevated when com-pared to background levels, the amountsof arsenic in bees at Sites 4 and 5 werebelow those generally associated with ar-senic toxicity.
Virtually nothing has been publishedabout cadmium toxicity to bees, although itis toxic to many other organisms. Our on-going studies indicate that cadmium is ap-proximately as toxic as arsenic (Cronn, un-published data). Lack of a correlationbetween cadmium levels in bee tissuesand observed colony responses resultedfrom somewhat higher values of cadmiumat Site 4 compared to Site 2 and does notmean that cadmium has no effect on beecolonies.
Based on our previous findings, cadmi-um is distributed somewhat differently fromarsenic around Puget Sound, probably be-cause arsenic was entering the atmos-
phere every day, whereas emissions near-ly 70 yr earlier produced most of thecadmium (Bromenshenk et al, 1985).
Synergistic or additive effects with ar-
senic may have occurred at Sites 1 and 3.In addition, the amount of cadmium at Site4, although considerably elevated abovebackground levels, could have been belowthe toxicity threshold.
The loss of wax from combs at 3 of thesites was not anticipated. Bees move waxwithin a colony to cap honey or to repaircomb (Dadant and Sons, 1975). Digitizedtracings and our visual observations indi-cate that bees may break down newlyformed wax comb. Whether this wax was
metabolized, used for capping, or eliminat-ed is unknown.
Based on all components measured,adult population size, as indicated by thenumber of frames covered by clusteredbees or the mass of bees, was the mostsensitive and least variable indicator of col-
ony effects. Frame counts were the easiesttest to conduct. Surplus honey may also in-dicate effects, but large within-site variabil-ity reduced our ability to distinguish statisti-cally significant changes in honey yield.Brood rearing, wax production, and foodstorage (pollen, nectar, and honey) infor-mation was useful for final interpretation. Italso provided data concerning the total dy-namics and energy flow within the colo-nies.
The results indicate that mini-coloniescan be used under field conditions to iden-
tify and quantify the effects of exposure toenvironmental contaminants such as
heavy metals. Low costs and manageabili-ty permit increased replication, which im-proves discernment of statistically signifi-cant responses.
ACKNOWLEDGMENTS
We wish to thank the beekeepers of PugetSound and residents of Vashon Island for theirassistance and support. Thanks also are due tothe Environmental Protection Agency Project Of-ficers E Preston and C Bishop for their assis-tance.
Résumé — Dynamique des populationsde petites colonies d’abeilles exposées
à des polluants d’origine industrielle.Des colonies d’abeilles de taille standardont déjà été utilisées pour contrôler des
polluants chimiques, principalement pourconnaître leur répartition à large échelle.Dans cette étude, ce sont de petites colo-nies, (4 cadres, environ 4 500 abeilles) quiont servi à étudier l’impact de la pollution.Cinquante nucléi ont été répartis en 5 sitesle long d’un gradient d’exposition à l’arse-nic et au cadmium, sur l’île de Vashon,dans l’état de Washington (fig 1a). Ces co-lonies ont subvenu elles-mêmes à leursbesoins et ont concurrencé avec succèsdes colonies standard dans l’exploitationdes ressources alimentaires. Au bout de40 jours, des différences statistiquementsignificatives sont apparues entre les sitesdans la masse et le nombre moyensd’abeilles (P ≤ 0,01, fig 1b), dans la pro-duction de miel (P ≤ 0,07, fig 2b) et dansles teneurs en arsenic et en cadmium desouvrières butineuses (P ≤ 0,001, fig 1d).Les populations d’abeilles étaient plus fai-bles d’environ 40% (fig 1 b) et la productionde miel d’environ 67% (fig 2b) sur le site leplus pollué à l’arsenic et au cadmium parrapport à celui qui l’était le moins. La taillede la population et les productions de mielétaient corrélées négativement (P ≤ 0,005)avec la teneur en arsenic des abeilles. Cesrésultats sont discutés en relation avec les
changements observés dans d’autres ca-ractéristiques des colonies, telles que le
stockage du miel et du nectar, les provi-sions de pollen, la production de cire, la
quantité de couvain operculé et non oper-culé, le poids de la ruche et le nombred’abeilles sur les rayons.
Apis mellifera / indicateur biologique /
pollution / dynamique des populations
Zusammenfassung — Populationsdyna-mik von Kleinvölkern bei Einwirkung in-dustrieller Schadstoffe. Bienenvölker von
Standardgröße wurden zur Erfassung che-mischer Schadstoffe benutzt, besondershinsichtlich deren großräumigen Vertei-
lung. In dieser Untersuchung hingegenwurden Kleinvölker (4-Waben-Ableger mitca 4500 Bienen) als Alternative zu Vollvöl-kern benutzt, um den Einfluß chemischerSchadstoffe zu erfassen. Fünfzig Kernvöl-ker wurden auf fünf Standorte entlangeines Expositionsgradienten für Arsen undCadmium auf Vashon Island, Washington(USA), aufgestellt (Abb 1 a). Diese Völkerkonnten sich selbst erhalten und konkur-rierten erfolgreich mit Vollvölkern bei denTrachtpflanzen. Nach 40 Tagen wurdenstatistisch signifikante Unterschiede zwi-schen den Standorten für die mittlereMasse und Zahl der Bienen (P ≤ 0,01, Abb1b) beobachtet, Honigertrag (P ≤ 0,07,Abb 2b) und den Arsen- und Cadmiumge-halt der Trachtbienen (P ≤ 0,001, Abb 1 d).Die Bienenmenge war am Standort mit derhöchsten Arsen- und Cadmiumbelastungum etwa 40% geringer als am Standort mitder geringsten Belastung, die Honigpro-duktion war sogar um 67% geringer. Popu-lationsgröße und Honigertrag waren mitdem Arsengehalt der Bienen negativ kor-reliert (P ≤ 0,005). Diese Ergebnissewerden in Zusammenhang mit Verände-rungen von anderen Volksmerkmalen dis-kutiert, wie Stapelung von Honig und
Nektar, Pollenvorräte, Wachserzeugung,Menge verdeckelter und unverdeckelterBrut, Volksgewicht und Anzahl der vonBienen bedeckten Waben.
Apis mellifera / Populationsdynamik /
Bioindikator / Schadstoff
REFERENCES
Anderson JF, Wojtas MA (1986) Honey bees(Hymenoptera: Apidae) contaminated with
pesticides and polychlorinated biphenyls.J Econ Entomol 79, 1200-1205
Bromenshenk JJ (1979) Monitoring environmen-tal materials and specimen banking using ter-restrial insects with particular reference to in-organic substances and pesticides. In: Moni-toring Environmental Materials and Speci-men Banking (Leupke NP, ed) Martinus Nij-hoff, The Hague
Bromenshenk JJ (1980) Accumulation andtransfer of fluoride and other trace elementsin honey bees near the Colstrip PowerPlants. In: The Bioenvironmental Impact of aCoal-Fired Power Plant; Fifth Interim Report.US Environmental Protection Agency, Cor-vallis Environ Res Lab, EPA-600/3-80-052,72-95
![Status of Beekeeping in Ethiopia- A Review · Beekeeping, also called apiculture, is management of honey bee colonies for pollination of crops and honey and other products [1]. It](https://static.documents.pub/doc/80x56/5e79e40d64fc9842da2e7688/status-of-beekeeping-in-ethiopia-a-review-beekeeping-also-called-apiculture-is.jpg)
