275Apidologie 35 (2004) 275282 INRA/DIB-AGIB/ EDP Sciences, 2004DOI: 10.1051/apido:2004014

Original article

Emergency queen rearing in honeybee colonies with brood of known age

Adam TOFILSKI*, Krystyna CZEKO SKA

Bee Research Department, Agricultural University, 29 Listopada 52 31-425 Krakow, Poland

(Received 25 April 2003; revised 14 July 2003; accepted 7 August 2003)

Abstract In four honeybee colonies, queens were isolated on empty combs for 8 consecutive days, so thatin every colony there were 8 combs containing brood of known age. Afterwards, the colonies weredequeened and the process of emergency queen rearing was observed. The average interval from egg layingto queen cell capping was 8.8 days and ranged from 7 to 12 days. The average interval from queen cellcapping to queen emergence was 7.2 days and ranged from 5 to 8 days. The whole development time fromegg laying to queen emergence was 15.7 days, ranging from 14 to 18 days. The age of brood at the momentof dequeening positively correlated with both the time of capping and the total queen development time.The average age of brood (at time of dequeening) around which queen cells were built was 3.0 days.However, higher proportions of queen cells with younger larvae were destroyed; in effect, the age of broodat dequeening from which queens emerged was 3.4 days.

Apis mellifera / honeybee / queen rearing / development time

N

1. INTRODUCTION

Honeybee (Apis mellifera L.) workers inqueenless colonies are able to rear queensfrom larvae primarily destined to be workers(Winston, 1979; Fell and Morse, 1984). Larvaefor queen rearing often are available in a rangeof ages. The reproductive quality (as defined byTarpy et al., 2000) of queens reared fromyounger larvae can be higher (Eckert, 1937;Boch and Jamieson, 1960; Weaver, 1957;Woyke, 1971; Tarpy et al., 2000). On the otherhand, rearing queens from older larvae canshorten the queenlessness period (Tarpy et al.,2000). Hatch et al. (1999) demonstrated thathoneybee workers rear emergency queens frombrood that at the time of dequeening are 1 to5 days old. However, their experimental setupdid not allow the age of brood in queen cells tobe determined directly. Instead they estimatedit based on the time of queen cell capping,assuming that queen cells are capped 8 days

after egg laying (Winston, 1987). This type ofestimation has often been used in studies ofemergency queen rearing (Winston, 1979; Felland Morse, 1984; Hatch et al., 1999; Schneiderand DeGrandi-Hoffman, 2002). The accuracyof this estimation has never been determinedeven though it is known that the time of queencell capping varies considerably. Jay (1963)reviewed the literature concerning the durationof queen development times, reporting thatqueen cells are capped between 7 and 9 daysafter egg laying and that the whole develop-ment time varies from 15 to 17 days. Genotypeis a major factor affecting queen developmenttime. There are marked differences betweensubspecies (Fletcher, 1978; Winston, 1979;DeGrandi-Hoffman et al., 1998) and some var-iation between colonies of the same subspecies(Visscher, 1986; Tarpy and Fletcher, 1998;Hatch et al., 1999). Even within the same colony,temperature (DeGrandi-Hoffman et al., 1993)and nutrition (Visscher, 1986) affect queen

* Corresponding author: rotofils@cyf-kr.edu.pl

276 A. Tofilski, K. Czeko skan

development time. Other factors that are relatedto temperature and nutrition, such as the positionof the queen cell in the nest (Visscher, 1986; Felland Morse, 1984; DeGrandi-Hoffman et al.,1993; Hatch et al., 1999), affect developmenttime as well.

Not all queen cells initiated by workers dur-ing an emergency queen rearing process suc-cessfully emerge. They can be destroyed bothbefore and after capping (Allen, 1956; Caronand Greve, 1979; Schneider et al., 2001, 2002).The destruction of capped queen cells can becarried out by both workers and newly emergedqueens (Allen, 1956; Fletcher, 1978; Caron andGreve, 1979; Gilley, 2001). Queens cut a smallhole in the side wall of the queen cell and some-times sting its occupant; workers cut a muchbigger hole in the side wall and instead of sting-ing they remove the contents of the queen cell(Allen, 1956; Fletcher, 1978; Caron and Greve,1979). Hatch et al. (1999) did not allow newlyemerged queens to walk freely in the nest anddestroy unemerged competitors in queen cells.Isolation of queen cells can affect the chancesof queens being reared from brood of differentages, and in consequence, affect the reproduc-tive quality of the only queen remaining in acolony when the process of emergency queenrearing is finished. Thus, to fully evaluate thefactors influencing queen rearing, it is neces-sary to examine queen replacement under nat-ural conditions in colonies containing brood ofknown age.

The purpose of this paper was to investigateemergency queen rearing in honeybee colonies.We had two main objectives. First, we deter-mined the variability of the different stages ofqueen development and the accuracy of esti-mation of ages of brood in queen cells based ontheir time of capping. Second, we measuredmore accurately than previous studies the agesof brood from which emergency queens areproduced. We examined the process of emer-gency queen rearing in colonies in which theage of young brood is known and newlyemerged queens have access to unemergedqueen cells.

2. MATERIALS AND METHODS

The experiment was carried out in June and July2000, and used four colonies of honeybees consid-

ered to be Apis mellifera carnica. The colonies wereof similar size, and each occupied two boxes con-taining 10 frames apiece. The frames were of thewielkopolski type locally used in Poland (width360 mm, height 260 mm). To obtain eggs of knownage, the queen from each colony was confined for24 hours on an empty frame of comb in a cage madeof wood and queen excluder. The caged queenswere placed in their original colonies in the centerof the bottom box. This procedure was repeated8 times. Every day the queens were isolated on newframes of empty comb. At the end of the procedureall colonies were dequeened. At that time in the bot-tom box of each hive there were 8 combs containingbrood ranging in age from 1-day old eggs to 5-dayold larvae. The other two frames in the bottom boxand the frames in the upper box contained broodolder than 8 days, honey and pollen stores. Thecombs with brood of different ages were placed in adifferent order in each colony (Fig. 1). Afterremoval of the queens, the colonies were examinedevery day at the same time of day until all queencells were either emerged or destroyed. Duringinspections the position of every queen cell wasmarked on acetate sheets, one sheet for each side ofeach frame. If a regular round opening was found atthe bottom of the queen cell it was categorized asemerged. If an opening was found in the side of thequeen cell it was categorized as destroyed. We callthe interval from egg laying to queen cell cappingthe precapping period, and the interval from queencell capping to queen emergence the postcappingperiod.

In the statistical analysis we combined the datafrom all four colonies because we were moreinterested in variation within the population than in

Figure 1. Arrangement of frames with brood ofdifferent ages in four experimental colonies.

Emergency queen rearing 277

differences between colonies. The times of queencell initiation, queen cell capping and queen emer-gence were analyzed in relation to both the time ofdequeening and the time of egg laying. Associa-tions between two variables (e.g., length of precap-ping period and distance from the bottom of thehive) were tested with Spearman's rank correlation.Differences between all continuous variables (e.g.,length of precapping period) and discontinuous var-iables (e.g., number of queen cells) were analyzedusing nonparametric tests: the Mann-Whitney testin the case of two groups and the Kruskal-Wallistest when there were more than two groups. Theproportions of queen cells destroyed were analyzedusing the G-test of independence (Sokal and Rohlf,1995). The G-test cannot be calculated for fre-quency equal to zero, so the numbers of emergedand destroyed queen cells were combined acrossages 6 to 8 days. All average values are reported as 1 SD.

3. RESULTS

3.1. General information

The numbers of initiated, capped and emergedqueen cells per colony were 32.7 4.79, 27.0 1.83 and 13.0 1.63 (N = 4), respectively. Thenumbers of queen cells on central and marginalframes were 19.0 2.45 and 13.7 2.63 (N =4), respectively. Those values differ signifi-cantly (Mann-Whitney test: U = 0.5, N1 = N2 =4, P = 0.028). The frame area covered by broodwas 1.55 0.55 dm2 (N = 32), ranging from 0.20to 2.76 dm2. There was no significant correla-tion between the frame area covered by broodand the number of queen cells on the frame(Spearmans rank correlation: rs = 0.208, N =32, P = 0.254). The brood area did not differbetween frames of different brood age (Kruskal-Wallis test: H = 5.18, df = 7, N = 32, P = 0.638),nor between central and marginal frames (Mann-Whitney test: U = 126, N1 = N2 = 16, P = 0.940).

3.2. Precapping period

The queen cell precapping period was 8.8 1.30 days (N = 108) and ranged from 7 to12 days (Fig. 2B). Queen cells from whichqueens emerged were capped at a slightly, butsignificantly earlier age (8.5 1.08 days; N = 52)than those that were destroyed (9.1 1.43 days;N = 56) (Mann-Whitney test: U = 1114, N1 =52, N2 = 56, P = 0.020; Fig. 2B). The precap-ping period of queen cells from which queensemerged positively correlated with the age of

brood at the time of dequeening (Spearmansrank correlation: rs = 0.555, N = 52, P < 0.001;Fig. 3A) and with the distance of the queen cellfrom the central frame (Spearmans rank cor-relation: rs = 0.317, N = 52, P = 0.022). The pre-capping period negatively correlated with thedistance of the queen cell from the bottom ofthe hive (Spearmans rank correlation: rs =0.304, N = 108, P = 0.001), but this relation-ship was not significant when only queen cellsfrom which queens emerged were analyzed(Spearmans rank correlation: rs = 0.141, N =52, P = 0.318). The precapping period nega-tively correlated with the postcapping period

Figure 2. Distribution of ages of brood aroundwhich queen cells were built (A), distribution ofages of brood at time of queen cell capping (B),distribution of ages of brood at time of queen celldestruction (C), and distribution of wholedevelopment times from egg laying to queenemergence (D) during emergency queen rearing.Bars represent averages + SD across four colonies.

278 A. Tofilski, K. Czeko skan

(Spearmans rank correlation: rs = 0.432, N =52, P = 0.001). The precapping period of queensreared from brood, which was younger at thetime of dequeening, tended to be shorter thanthe precapping period of queens reared fromolder brood. The queen cells in the center of thenest tended to be capped after a shorter periodof time than the queen cells at the nest periphery.

3.3. Postcapping period

The postcapping period was 7.2 0.78 days(N = 52) and ranged from 5 to 8 days (Fig. 3B).The postcapping period negatively correlatedwith the distance of the queen cell from thecentral frame (Spearmans rank correlation:rs = 0.311, N = 52, P = 0.025). There was nosignificant relationship between the postcap-ping period and the age of brood at the time of

dequeening (Spearmans rank correlation: rs =0.082, N = 52, P = 0.563; Fig. 3B) and thedistance of the queen cell from the bottom ofthe hive (Spearmans rank correlation: rs =0.172, N = 32, P = 0.222). The postcappingperiod of queens reared in the center of thenest tended to be longer than the postcappingperiod of queens reared at the nest peripherybut was not affected by the age of brood at thetime of dequeening.

3.4. Total development time

Queen emergence occurred between 14 and18 days after the egg was laid, with an averageof 15.7 0.88 days (N = 52; Fig. 2D). The totaldevelopment time positively correlated withthe age of brood at the time of dequeening(Spearmans rank correlation: rs = 0.561, N =52, P < 0.001; Fig. 3C). There was no signifi-cant relationship between the total develop-ment time and the distance of the queen cellfrom the central frame (Spearmans rank cor-relation: rs = 0.118, N = 52, P = 0.406) and thedistance of the queen cell from the bottom ofthe hive (Spearmans rank correlation: rs =0.017, N = 52, P = 0.904). The total develop-ment time of queens reared from brood that wasyounger at the time of dequeening tended to beshorter than the total development time of queensreared form older brood, but did not depend onthe position of the queen cells in the nest.

3.5. Initiation and destruction of queen cells

The queen cells were initiated around broodaged between 3 and 11 days; the average ageof brood used to initiate queen cells was 5.9 1.90 days (N = 131; Fig. 2A). Eggs were neverobserved inside queen cells. Queen cells fromwhich queens emerged and those destroyedeither before or after capping did not differ inthe age of brood from which they were con-structed (Mann-Whitney test: U = 1732, N1 =52, N2 = 79, P = 0.123; Fig. 2A). Most of thequeen cells (60.3%) were destroyed, 17.6% ofthem before capping and 42.7% after capping(Fig. 2C). The queen cells were destroyedbetween the 5th and 18th days of brooddevelopment, and the average age of broodat the time of queen cell destruction was13.0 3.48 days (N = 79; Fig. 2C). A smaller

Figure 3. Relationship between age of brood attime of dequeening and different stages of queendevelopment. Only data from emerged queen cellswere included. Sizes of squares correspond tonumber of observations.

Emergency queen rearing 279

proportion of queen cells were destroyed oncentral frames (51.3%) than on marginalframes (72.7%; G-test of independence: G =6.17, df = 1, P = 0.013).

3.6. Use of brood of different ages

The age (at time of dequeening) of broodaround which queen cells were built and fromwhich queens emerged was 3.0 1.40 (N = 131)and 3.4 1.27 (N = 52) days, respectively(Fig. 4). Queen cells from which queensemerged were initiated using brood that wasolder at the time of dequeening (3.4 1.27 days;N = 52) than the brood of those destroyed eitherbefore or after capping (2.7 1.40 days; N =79) (Mann-Whitney test: U = 1327, N1 = 52,N2 = 79, P < 0.001; Fig. 4). The age of broodat the time of dequeening significantly affectedthe number of queen cells constructed using thebrood (Kruskal-Wallis test: H = 21.2, df = 7,N = 32, P < 0.004; Fig. 4). The greatest numberof queen cells were constructed over brood3 days old at the time of dequeening (Fig. 4).The age of brood at the time of dequeening alsoaffected the proportion of queen cells destroyed(G-test of independence: G = 17.0, df = 5, P =0.005; Fig. 4). The smallest proportion of queencells were destroyed on frames with brood4 days old at the time of dequeening (Fig. 4).

3.7. Emergency queen rearingin relation to time of dequeening

Queen cells were initiated between the 1stand 9th days after dequeening (Fig. 5A). In twoof the colonies the first queen cells wereobserved the day after dequeening, and in theother two colonies the second day after

dequeening. Queen cells from which queensemerged were initiated significantly earlierthan those destroyed either before or after cap-ping (Mann-Whitney test: U = 1372, N1 = 52,N2 = 79, P = 0.001; Fig. 5A). The distributionof times at which queen cells were initiated wasbimodal, with the first mode on the 2nd3rd dayafter dequeening and the second mode onthe 9th day after dequeening (Fig. 5A). Thequeen cells were capped between the 4th and11th days after dequeening, and the peak ofqueen cell capping occurred the 6th7th dayafter dequeening (Fig. 5B). The queen cellsfrom which queens emerged were sealed sig-nificantly earlier than those destroyed after cap-ping (Mann-Whitney test: U = 752, N1 = 52,N2 = 56, P < 0.001; Fig. 5B). Queens emerged

Figure 4. Distribution of ages of brood (at time ofdequeening) from which queen cells were built.Bars represent averages + SD across four colonies.

Figure 5. Timing of rebuilding of worker cells intoqueen cells (A), capping of queen cells (B),destruction of queen cells (C), and queenemergence (D) during emergency queen rearing.Bars represent averages + SD across four colonies.

280 A. Tofilski, K. Czeko skan

from the queen cells between the 11th and16th days after dequeening, and the peak ofqueen emergence occurred the 13th day afterdequeening (Fig. 5D). Queens reared frombrood older at the time of dequeening tendto emerge earlier than queens reared fromyounger brood (Spearmans rank correlation:rs = 0.758, N = 52, P < 0.001). Destruction ofqueen cells was observed between the 3rd and16th days after dequeening (Fig. 5C). The dis-tribution of times at which queen cells weredestroyed was bimodal, with the first mode onthe 4th6th day after dequeening and the secondon the 12th day after dequeening (Fig. 5C).

4. DISCUSSION

Our data show that both the precappingperiod and the whole development time ofemergency queens increase with the age ofbrood from which the queens were reared (Fig. 3).Thus, estimations of brood age in emergencyqueen cells based on time of capping (Winston,1979; Fell and Morse, 1984; Hatch et al., 1999;Schneider and DeGrandi-Hoffman, 2002) can-not be very accurate. Another source of inac-curacy of estimations is the correlation of theprecapping period with the position of thequeen cell in the nest (Visscher, 1986; Fell andMorse, 1984; DeGrandi-Hoffman et al., 1993;Hatch et al., 1999). Both the distance from thecentral frame and the distance from the bottomof the nest affected the length of the precappingperiod in this study. The differences in thelength of the precapping period of brood inqueen cells positioned in different parts of thenest are compensated by the length of the post-capping period. In consequence the wholedevelopment time does not depend on the posi-tion of the queen cell in the nest. Bienefeld(1996) observed a similar negative correlationbetween the precapping and postcapping peri-ods in honeybee workers.

Estimating age of brood in queen cells onthe assumption that the precapping period isalways 8 days often results in underestimationsof brood age. In the majority of cases the under-estimation is low compared to the error of oneday accepted by most studies of emergencyqueen rearing (Hatch et al., 1999; Schneiderand DeGrandi-Hoffman, 2002). For the aver-age age of brood used to produce queens in thisstudy (2.96 days; Fig. 4), the difference between

the estimated and actual precapping period (cal-culated using linear regression coefficients;Fig. 3A) was 0.80 days. However, some of thequeen cells were capped 12 days after egg lying(Fig. 2B). In those rare (3%) cases the error ofestimation of the age of brood in queen cells canreach four days. During swarming, all queencells are initiated with newly laid eggs, so broodage estimations based on the time of capping canthen be much more accurate, although this needsto be verified experimentally. Other factoraffecting the accuracy of the estimation will bethe position of the queen cell in the nest. Ourresults suggest that the whole developmenttime, instead of the time of capping, should beused for estimation because it does not dependon the position of the queen cell in the nest.

We studied emergency queen rearing undernatural conditions and did not isolate thecapped queen cells. This means that the time ofemergence of queens was affected not only bythe age of brood (at time of dequeening) and theposition of queen cells in the nest but also byother factors that can influence queen emer-gence and which may differ between emer-gency queen rearing and swarming. Workersstanding on queen cells often perform a vibra-tion signal, which consists of dorsoventralvibration of their body (Allen, 1959; Painter-Kurt and Schneider, 1998). It has been sug-gested that the vibration signal affects timing ofqueen emergence (Fletcher, 1978; Bruinsmaet al., 1981; but see Grooters, 1987; Schneideret al., 2001) and inhibits interactions betweenqueens (Fletcher, 1978; Schneider, 1990, 1991).Moreover, emerged queens produce a series ofpulsed sounds called piping which can delay theemergence of other queens (Grooters, 1987).Piping is more common during swarming thanduring emergency queen rearing (personal obser-vations). A factor probably limited to swarmingperiods is imprisonment of young queens inqueen cells (Fletcher, 1978; Bruinsma et al.,1981). The mentioned factors may influence thetiming of queen emergence and thus may haveaffected the positive relationship between age ofbrood (at time of dequeening) and the totaldevelopment time. Other experimental setupsand particularly isolation of capped queen cellsin an incubator can influence the length of dif-ferent stages of queens development. This hasto be taken into account when the data presentedhere are compared with results of other studies.

Emergency queen rearing 281

The average lengths of different stages ofhoneybee queen development from our studyagree with those from other studies (Jay, 1963;Winston, 1987), but the variation was greaterin our experiment (Fig. 2). The large variationcan be explained partly by differences in meas-urement accuracy, which in our study wasaffected by the length of time between consec-utive queen cell inspections (24 h) and also bythe length of time provided for queens to layeggs in a single frame of comb (24 h). It is dif-ficult to obtain higher accuracy of brood agemeasurements in a full-size colony, becausefrequent inspections disturb it and can affect theresults.

We showed that workers initiate queen cellsusing brood in a wide range of ages at dequeen-ing, but the greatest number of queen cells wereproduced around brood that at the time ofdequeening were 3 days old (Fig. 4). Hatchet al. (1999) reported a similar pattern of queencell initiation, but the range of ages used to pro-duce queen cells was narrower in their study.This discrepancy is probably due to inaccurateestimates of brood age, based on the time ofqueen cell capping. Even though newlyemerged queens are able to destroy other queencells, in our experiment a considerable numberof queens emerged in each of the colonies.Either workers protected the queen cells fromdestruction (Gilley, 2001) or the early emergedqueens could not find and destroy the remainingqueen cells before the queens emerged fromthem. Because queen cells with younger broodare more often destroyed, the average age ofbrood (at time of dequeening) from whichqueens emerged was higher than the averageage of brood (at time of dequeening) aroundwhich queen cells were built (Fig. 4). We didnot observe queen cell destruction in progress;another experiment is needed to determinewhether the queen cells with younger broodwere destroyed by newly emerged queens orworkers. This problem needs to be addressed sothat the mechanisms by which the quality ofemergency queens is controlled in honeybeecolonies can be better understood.

ACKNOWLEDGMENTS

We thank Michael Jacobs, Luis Medina Medina,Stanley Schneider and two anonymous referees forhelpful comments on earlier versions of this paper.

Rsum Production de reines de remplacementdans des colonies dabeilles (Apis mellifera) dontlge du couvain est connu. Dans les coloniesdabeilles orphelines les ouvrires sont capablesdlever des reines de remplacement partir de lar-ves qui taient destines devenir des ouvrires.Dans les tudes prcdentes sur la production desreines de remplacement, lge du couvain choisi parles ouvrires tait estim daprs le moment oavait eu lieu loperculation de la cellule royale. Lebut de cette tude tait (i) de dterminer la prcisionde lestimation de lge du couvain dans les cellulesroyales et (ii) de mesurer lge des larves partirdesquelles les reines de remplacement taient le-ves. Dans quatre colonies on a isol les reines surun rayon vide chaque jour durant 8 j conscutifs. Ily avait donc dans chaque colonie 8 rayons renfer-mant du couvain dge connu. Aprs avoir t lesreines, on a examin journellement et not la posi-tion et la condition de chaque cellule royale.Lintervalle entre la ponte et loperculation de lacellule royale tait comprise entre 7 et 12 j, avec unemoyenne de 8,8 j (Fig. 2B). La reine a clos 7,2 jplus tard, en moyenne (Fig. 3B). La corrlationentre ces deux dures de dveloppement tait nga-tive. La dure totale de dveloppement de la ponte lclosion de la reine tait comprise entre 14 et 18 j,avec une moyenne de 15,7 j (Fig. 2D). Lge du cou-vain au moment de lorphelinage tait corrl posi-tivement avec la dure de la priode pr-opercula-tion et la dure totale de dveloppement (Fig. 4). Aumoment de lorphelinage, le couvain autour duquelles cellules royales taient construites avait un gemoyen de 3,0 j. Pourtant les cellules royales poss-dant des larves plus jeunes ont t dtruites en plusgrand nombre ; au moment de lorphelinage, lgedes larves do sont sorties les reines tait de 3,4 j(Fig. 4). Ce rsultat montre que lestimation delge du couvain base sur la priode de pr-opercu-lation ne peut pas tre trs prcise.

Apis mellifera / levage de reines / dure dedveloppement

Zusammenfassung Erzeugung von Nachschaf-fungskniginnen in Bienenvlkern mit Brutbekannten Alters. Honigbienen in weisellosen Vl-kern knnen Nachschaffungskniginnen aus Larvennachziehen, diese werden von den Arbeiterinnenbestimmt. In frheren Untersuchungen wurde dasAlter der von den Arbeiterinnen zur Kniginnener-zeugung ausgesuchten Larven aus dem Zeitpunktder Verdeckelung der Kniginnenzellen geschtzt.Das Ziel dieser Untersuchung war, das Alter der Brutzu messen, aus der die Nachschaffungszellen erzeugtwerden und hieran weiterhin die Genauigkeit derbisherigen Altersschtzung zu berprfen. In vierBienenvlkern wurden die Kniginnen an acht aufei-nanderfolgenden Tagen auf leere Waben gesperrt,hierdurch befanden sich in jedem Volk 8 Wabenmit Brut bekannten Alters. Nach Entfernung der

282 A. Tofilski, K. Czeko skan

Kniginnen wurden die Vlker jeden Tag untersuchtund die Lage und der Zustand jeder entstehendenNachschaffungszelle notiert. Das mittlere Zeitintervallzwischen der Eilage und der Verdeckelung der Kni-ginnenzellen betrug 8,8 Tage, es schwankte zwi-schen 7 und 12 Tagen (Abb. 2B). Danach dauerte esim Mittel weitere 7,2 Tage bis die Kniginnenschlpften (Abb. 3B). Diese zwei Entwicklungs-zeiten korrelierten negativ und die gesamte Entwick-lungszeit von der Eilage bis zum Schlupf der Kni-ginnen betrug 15,7 Tage, es schwankte zwischen 14und 18 Tagen (Abb. 2D). Das Alter der Brut zum Zeit-punkt der Entweiselung korrelierte positiv sowohlmit der Dauer der Zeit vor der Verdeckelung als auchmit der gesamten Entwicklungsdauer (Abb. 4). ZumZeitpunkt der Entweiselung betrug das Alter der Brut,aus der die Nachschaffungszellen erzeugt wurden,3,0 Tage. Allerdings wurden grere Anzahlen deraus jngerer Brut erzeugten Kniginnenzellen zer-strt, hierdurch betrug das Alter der Larven aus denentatschlich Kniginnen schlpften, beim Entweiselnder Vlker 3,4 Tage (Abb. 4). Die Ergebnisse zeigen,dass die auf der Zeit bis zum Verdeckeln der Zellenberuhenden Altersschtzungen der Larven nicht sehrgenau sein knnen.

Apis mellifera / Honigbienen / Kniginnenerzeu-gung / Entwicklungsdauer

REFERENCES

Allen M.D. (1956) The behavior of honeybees prepa-ring to swarm, Brit. J. Anim. Behav. 4, 1422.

Allen M.D. (1959) The occurrence and possiblesignificance of the shaking of honeybee queensby workers, Anim. Behav. 7, 6669.

Bienefeld K. (1996) Factors affecting duration of thepostcapping period in brood of the honey bee, J.Apic. Res. 35, 1117.

Boch R., Jamieson C.A. (1960) Relation of bodyweight to fecundity in queen honeybees, Can.Entomol. 92, 700701.

Bruinsma O., Kruijt J.P., van Dusseldorp W. (1981)Delay of emergence of honey bee queens inresponse to tooting sounds, Proc. K. Ned. Akad.van Wet., Series C 84, 381387.

Caron D.M., Greve C.W. (1979) Destruction of queencells placed in queenright Apis mellifera colonies,Ann. Entomol. Soc. Am. 72, 405407.

DeGrandi-Hoffman G., Spivak M., Martin J.H.(1993) Role of thermoregulation by nestmates onthe development time of honey bee (Hymenop-tera: Apidae) queens, Ann. Entomol. Soc. Am.86, 165172.

DeGrandi-Hoffman G., Watkins J.C., Collins A.M.,Loper G.M., Martin J.H., Arias M.C., SheppardW.S. (1998) Queen developmental time as afactor in the Africanization of European honeybee (Hymenoptera: Apidae) populations, Ann.Entomol. Soc. Am. 91, 5258.

Eckert J.E. (1937) Relation of size to fecundity inqueen honeybees, J. Econ. Entomol. 30, 646648.

Fell R.D., Morse R.A. (1984) Emergency queen cellproduction in the honey bee colony, Insectes Soc.31, 221237.

Fletcher D.J.C. (1978) The African bee, Apis mellif-era adonsoni, in Africa, Annu. Rev. Entomol.23,151171.

Gilley D.C. (2001) The behavior of honey bees (Apismellifera ligustica) during queen duels, Ethology107, 601622.

Grooters H.J. (1987) Influences of queen piping andworker behaviour on the timing of emergence ofhoney bee queens, Insectes Soc. 34, 181193.

Hatch S., Tarpy D.R., Fletcher D.J.C. (1999) Workerregulation of emergency queen rearing in honeybee colonies and the resultant variation in queenquality, Insectes Soc. 46, 372377.

Jay S.C. (1963) The development of honeybee in theircells, J. Apic. Res. 2, 117134.

Painter-Kurt S., Schneider S.S. (1998) Age and behav-iour of honey bees, Apis mellifera (Hymenoptera:Apidae), that perform vibration signals on queensand queen cells, Ethology 104, 475485.

Sokal R.R., Rohlf F.J. (1995) Biometry, W.H.Freeman, New York.

Schneider S.S. (1990) Queen behaviour and workerqueen interactions in absconding and swarmingcolonies of the African honey bee, Apis melliferascutellata (Hymenoptera: Apidae), J. Kans. Ento-mol. Soc. 63, 179186.

Schneider S.S. (1991) Modulation of queen activityby the vibration dance in swarming colonies ofthe African honey bee, Apis mellifera scutellata(Hymenoptera: Apidae), J. Kans. Entomol. Soc.64, 269278.

Schneider S.S., DeGrandi-Hoffman G. (2002) Theinfluence of worker behavior and paternity onthe development and emergence of honey beequeens, Insectes Soc. 49, 306314.

Schneider S.S., Painter-Kurt S., DeGrandi-HoffmanG. (2001) The role of the vibration signal duringqueen competition in colonies of the honeybees,Apis mellifera, Anim. Behav. 61, 11731180.

Tarpy D.R., Fletcher D.J.C. (1998) Effects of related-ness on queen competition within honey bee col-onies, Anim. Behav. 55, 537543.

Tarpy D.R., Hatch S., Fletcher D.J.C (2000) Theinfluence of queen age and quality during queenreplacement in honeybee colonies, Anim. Behav.59, 97101.

Visscher P.K. (1986) Effect of location within the neston acceptance of queen cells in honeybeecolonies, J. Apic. Res. 25, 154157.

Weaver N. (1957) Effects of larval age on dimorphicdifferentiation of the female honeybee, Ann.Entomol. Soc. Am. 50, 283294.

Winston M.L. (1979) Events following queenremoval in colonies of Africanized honeybees inSouth America, Insectes Soc. 26, 373381.

Winston M.L. (1987) The biology of the honey bee,Harvard University Press, Cambridge, London.

Woyke J. (1971) Correlations between the age atwhich honeybee brood was grafted, characteris-tics of the resultant queens, and results of insemi-nation, J. Apic. Res. 10, 4555.