Original article

Diurnal and seasonal variation in defensive behaviorof African bees Apis mellifera adansonii in Ghana

J Woyke

FAO Representation, Accra and Technology Consultancy Centre of Kumasi University, Ghana

(Received 1 September 1991; accepted 27 April 1992)

Summary — The defense behavior of an African bee colony was studied in order to determine rec-ommendations for handling A m adansonii bees (when and how) without being stung. The number ofstings left in a leather ball suspended 50 cm in front of the hive entrance for 1 min was thereforecounted as well as the number of workers returning to the hive over a 5-min period. A typical patternof diurnal defensive behavior was observed. During the non-productive period, ie the summer rainyseason, one peak of defensive behavior was observed in the morning. During the period of honeyflow, ie the winter dry season, 2 peaks occurred, one early in the morning and the other late in theevening. A highly positive correlation existed between flight activity and defensive behavior. Thebees stung less during periods of low flight activity. Bees in front of the entrance defended the col-ony more than those flying out from the combs. During the period of honey flow (the winter dry sea-son), the bees stung about 4 times more frequently than during the summer rainy season. The ratioof the number of stinging bees to that of flying bees is a better indicator for comparison of potentialdefensive behavior of different colonies than that of the absolute number of stings. It is recommend-ed that handling of A m adansonii take place during periods of low flight activity.

Apis mellifera adansonii / defensive behavior / activity rhythm / flight activity / Ghana

INTRODUCTION

The highly defensive behavior of Africanbees is well known. It is therefore impor-tant to determine the periods during whichthe bees are more aggressive and thosewhen they are less aggressive. On thisbasis, suggestions can be made aboutwhen and how to work with the bees. The

defensive behavior of honey bees in thefield has been investigated several times.

To further examine this aspect, the num-ber of stings left in a dark leather ball sus-pended in front of the hive entrance wascounted (Free, 1961; Stort, 1970, 1974,1975; Brandenburgo et al, 1977; Collins etal, 1982; Collins and Kubasek, 1982; Villa,1988). The colony defense behavior of Af-ricanized and European honeybees wasthen compared. A model of honeybee de-fensive behavior, involving 4 steps, hasbeen presented by Collins et al (1980).Experiences with Africanized bees under

* Present address: Bee Division, Agriculture University, SGGW, 166 Nowoursynowska, 02-766 War-saw, Poland

very different conditions at sites in Braziland Poland have been described by Woyke(1973). The latter author has pointed out(1989a) that a correlation exists betweenflight activity and defensive behavior.

During the daily work with bees in Gha-na, 3 distinct phenomena concerning thedefensive behavior of Apis mellifera adan-sonii; were observed, ie: 1), wide variationin defensive behavior during the day; 2)when smoke was directed through the hiveentrance into the bee nest, worker bees

escaped in great numbers and started tosting the beekeeper. The same phenome-non occured when the hive was openedand smoke was directed between the

combs; 3), the bees exhibited non-

agressive behavior during the non-

productive period, ie the summer rainyseason, and were very aggressive duringthe winter dry season, ie the period of hon-ey flow. Individual discussions as well asthose conducted at beekeepers’ meetingsshowed that these phenomena had beennoticed and were familiar to many Ghan-ian beekeepers. Quantitative investiga-tions were therefore conducted to deter-mine whether the daily practical observa-tions could be scientifically confirmed.

MATERIALS AND METHODS

The study took place in Ghana over a 2-monthperiod (26/5/1986-26/7/1986) during the summerrainy - non-productive season, and over a 3-month period (23/11/1986-21/2/1987) during thewinter dry-honey flow season. The authorworked with the bees daily for 10-12 h. At differ-ent times, 20-30 bee colonies were present atthe university apiary in Kumasi. However, bees inother apiaries in north, central and south Ghanawere also studied (Woyke, 1989a,b, 1992). De-tailed investigations were conducted in Kumasi incentral Ghana. The defensive behavior of 4 Apismellifera adansonii bee colonies was studied.The colonies were kept in Kenya top bar hivesplaced on individual stands several meters apart.

Investigations were conducted every hour from6.00 to 18.00 h, ie 13 times a day. The number ofworker bees returning to the hive over a 5-minperiod was counted (Woyke, 1992). These num-bers were divided by 5 to obtain results relating toa 1-min period, which were used for further calcu-lations and comparisons. The defensive behaviorof worker bees was then investigated. To thisend, a black leather ball (diameter 7 cm) wasturned with the rough side outside. Before theball was used for tests, it was placed close to theentrance of a hive (not used in the experiment) tobe stung by dozens of workers, so that it becameimpregnated with the smell of venom. For the testthe ball was suspended on a thread 50 cm infront of the hive entrance, and was set in oscilla-

tory motion for 1 min. Stings that had lodged inthe ball were then counted and removed. On

February 8th, 2 additional tests were conducted.Immediately after the first test was completed,the ball was suspended for 1 min 25 cm abovethe top bars of the same colony, while 2 beewayswere opened between 3 combs. After the stingshad been removed and counted, 2 puffs ofsmoke originating from buming dry leaves weredirected between the 2 opened beeways, and thetest was repeated. At the same time, tempera-ture, RH and degree of cloudiness were noted.

The area occupied by brood was measuredin 3 colonies at the time of the investigations.Measurements of natural cells showed, that5.067 cells occupied an area of 1 cm2. Threecolonies were weighed at weekly intervals.

Analysis of variance was performed. New-man-Keuls multiple comparison test was usedto determine significant differences between

means. χ2 test was used to compare frequencydistributions.

RESULTS

Diurnal pattern of defensive behavior

A wide variation in diurnal defensive be-havior was recorded during the day in dif-ferent bee colonies: from 2-112 stingsover a 1-min period in colony No 5, to 33-144 stings/min in colony No 3 (both on

June 23, 1986), or 8-248 stings/min in col-ony No 4 (January 13, 1987).A characteristic pattern of diurnal defen-

sive behavior was noted. During the dearthperiod, ie the rainy season (June, 23), de-fensive behavior of colony No 4 rose from18 stings/min at 6 h to a peak of 62 stings/min at 9.00 h (fig 1). Later, the defensiveactivity declined. Some variation occurredduring the day. Low defensive behaviorwas observed at midday (11 stings/min at13 h), with the lowest activity recorded inthe evening (5-6 stings/min at 17.00 and18.00 h respectively).

Considerable changes occurred duringthe period of honey flow, ie the winter dryseason (January 13; fig 2). Three peaks ofcolony defense were recorded. One earlymorning peak of 107 stings/min occurredat 6 h, 30 min before sunrise. Then, defen-sive activity dropped to 11 stings/min at

8 h, and reached the highest peak of 248stings/min at 10 h. Defensive behavior

again dropped to 8 stings/min at 14 h, butrose to 163 stings/min at 18 h just beforesunset. About 1 month later (fig 3), only 2

peaks of defensive activity were recorded,one early in the morning with 141 stings/min and the other late in the evening at18 h, with 144 stings/min just at sunset.

During the rest of the day the bees exhibit-ed gentle behavior. The evening peak ofcolony defense was characteristic for bothdays during the honey flow season.

Correlation between defensive behaviorand flight activity

Figures 1 and 2 as well as other recordeddata show that stinging intensity was relat-ed to flight activity. Statistical calculationsrevealed a significant correlation betweenstinging and flight activity during the day.The correlation coefficients for colony No 4were as follows: (df = 11); r = 0.96 (June23) or r = 0.66 (June 24) during the dearthperiod-summer rainy season; and r= 0.71(Jan 13) and r = 0.96 (February 8) duringthe honey flow period-winter dry season.The equation for the regression line pre-sented in figure 4 is y = -3.26 + 0.24 x (y=No stinging bees; x = No of returning for-

agers/min). The slope as well as the inter-cept for the other days were different.

The ratio between the number of sting-ing bees and returning workers was calcu-lated. For easier presentation and assess-ment, it was multiplied by 100 and calledthe percentage. This ratio did not remainconstant during the day. Bee colony No 4showed a range of 12-41% stinging beesduring the dearth period-summer rainyseason (June 23; fig 1), to 24-212% dur-ing the honey flow period-winter dry sea-son (January 13; fig 2). Thus sometimesonly a low percentage of flying bees en-gaged in stinging activity, but sometimestwice as many bees were stinging as re-turning to the hive. The other stinging beeswere workers flying out of the hive and theguard bees. Figures 1 and 2 show that in

general the ratio of stinging bees was high-est at midday.

Taken together, the results justify a gen-eral prediction: low relative flight activity in-dicates low defensive behavior, the op-posite behavior pattern to that expectedwith European bees.

Defensive behavior of bees in frontof the entrance and of bees flying outof beeways between combs

Specific tests (fig 3) showed that between6-17 h, 7-141 workers stung the ball sus-pended 50 cm in front of the entrance;however, only 0-6 workers stung a ball

suspended 25 cm above the top bars of 2opened beeways. Tests were also carriedout with a ball suspended 50 cm above thetop bars. However, the bees practically didnot sting the ball that was suspended sohigh. Thus, bees in front the entrance de-fended the colony more vigorously thanthose flying out of beeways between thecombs. More surprising was the observa-

tion that as many as 2-60 bees from bee-

ways between combs stung the ball afterthey had been smoked out. An especiallyvigorous reaction to smoking was ob-

served after the first treatments at 6 and7 h. The bees were also particularly defen-sive at 18 h during the period of maximalflight activity. They flew out of the hive notonly through the entrance and the 2

opened beeways between combs, but alsothrough all cracks and openings. They at-tacked everything so vigorously that thesmoking test could not be conducted at

that time. Between 6-17 h, a total of 428workers stung the ball in front of the en-

trance, 16 the ball suspended 25 cmabove the top bars of 2 opened beewaysand 149 after the bees had been smokedout with 2 puffs of smoke. The ball in frontof the entrance was stung on average (n =12; x ± SEM) 35.7 ± 11.1 times per test,the ball above the opened beeways 1.3 ±0.5 times before smoking and 12.4 ± 5.0times after the bees had been smoked out.Paired t-test showed a statistically signifi-cant difference between the 2 latter

means, t = 2.37; df = 11; P = 0.037. Thussmoking significantly increased the defen-sive behavior of worker bees. Workers

stung the ball on the average 10 times

more after than before smoking.Correlation coefficients between the

number of foragers and the number of

stingers in front of the entrance, abovecombs before and after smoking were: r =0.93, r = 0.86 and r = 0.87 respectively.Regression between these characters wasy = 7.87 ± 0.47f, y = 0.26 + 0.02f and y =0.64 + 0.20f g respectively (f = No forag-ers). The values of the intercept and theslope for number of stinging bees after thebees had been smoked out were betweenthose for the number of stinging bees in

front of the entrance and the numberabove the combs before the bees hadbeen smoked out. All these results confirm

the daily practical observations accordingto which bees started to sting after thesmoke had been directed through the en-trance into the bee nest. It can be conclud-ed that the smoke irritated the bees and in-creased their defensive reaction.

Defensive behavior in different seasons

Defensive behavior was investigated in

summer and winter. The temperature at

14 h on January 13 was 4 °C higher, andon February 8, 7 °C higher than in sum-

mer, although the early morning tempera-ture on January 13 was the lowest (fig 5).June falls during the non-productive peri-od-summer rainy season. January andFebruary fall during the honey flow period-winter dry season. The bee colonies lost1-2 kg/colony/week in June and in Janu-ary gained 1-1.5 kg/colony/week. Thenumber of brood combs in colony No 4 on

June 26, January 15 and February 5 was10, 8 and 7 (combs) respectively. Thebees covered 4—5 more combs.

Figure 1 shows that one morning peakof defensive behavior was observed onJune 23, during the non-productive peri-od-summer rainy season. A similar pat-tern in the same colony was noted thenext day. Figures 2 and 3 show that at

least 2 peaks, one early in the morningand the other late in the evening, were ob-served during the honey flow period-winter dry season. Thus the pattern ofdiurnal defensive behavior of the same

colony did not remain constant. It maychange drastically during different periodsof the year. The 2 patterns of colony de-fense agreed very well with the 2 patternsof flight activity recorded by Woyke (1992)during 10 different days of investigation inboth seasons in Ghana.

The average number of flights/min dur-ing the entire day, as well as the average

number of stings/min during the day wascalculated (fig 6). χ2-Test showed that fre-quency distribution of the average numberof foragers on different days differed signif-icantly from that of the number of broodcombs present in the colonies at the sametime: χ2calc = 12.5 > χ2tab 0.05 = 7.81, P =0.006 (df = 3). Thus the amount of brood(noted above) did not show a correlationwith the number of foragers.

Statistical calculations did not show sig-nificant differences in the average numberof foragers in different seasons, as pres-ented in figure 6. However, the χ2-testshowed that the averages differed from the1:1:1:1 frequency distribution: χ2calc =

12.78 > χ2tab 0.05 = 7.81, P = 0.006 (df =3). No significant differences were foundbetween both averages of the number of

stinging bees in each season. The averagenumber of stinging bees on January 13was significantly higher than both summeraverages. χ2-Test showed that the fre-

quency distribution of the average num-bers of stinging bees on both summer

days and on February 8 (21, 21 and 44respectively) differed significantly from the1:1:1 distribution: χ2calc = 12.30 > χ2tab 0.05= 5.99, P = 0.02 (df = 2). The ratio (%) ofstinging bees to foragers did not differ with-in the same season; however, it was signif-icantly higher on both days during the hon-ey flow season than during the non-

productive period. Since the ratio amount-ed to 20-22% in summer and 80-81% in

winter, this indicates that the defensive be-havior of the same colony on the days ofthe investigation was about 4 times higherduring the honey flow season than duringthe non-productive season. Thus the ratioof stinging bees to foragers does not re-main constant throughout the year.

Comparison of defensive behaviorin 4 bee colonies

Figure 7 shows that differences exist in thediurnal defensive behavior of different beecolonies. The pattern of defensive behav-

ior in 3 colonies was similar. The morningpeak of the 4th colony, No 3, occurred lat-er (June 23). The next day the pattern wassimilar, but the peaks of the same coloniesshifted as a function of time in comparisonto the other colonies. The variation in de-fensive behavior between different col-

onies was statistically significant for bothdays.

The differences in the mean numbers of

stings left in the test ball on June 23 werefound to be statistically significant only be-tween 1 colony and the 3 others (table I).The next day, differences between 3 col-onies were found to be significant. Aftermeans for both days had been calculated,significant differences were only found be-tween 2 colonies (table I): colony No 3 ex-hibited the most defensive, and No 4 theleast defensive behavior.

Ratio of stinging bees to foragersin 4 colonies

A high correlation was found to exist be-tween flight activity and stinging in colonyNo 4. Flight activity of 4 bee colonies wastherefore investigated at 10, 11, 12, 14 and15 h. The relationship between average

flight activity and colony defense was cal-culated (fig 8) and a statistical calculationwas carried out.

At the time of the investigation, 11, 10 and 8 brood combs containing 31.0, 39.9and 33.8 thousand brood cells were foundin colonies Nos 3, 4 and 5 respectively.The worker bees covered 4 more combs.The brood in colony No 6 was not meas-ured. χ2-Test showed that the distributionof the average number of foragers fromvarious colonies (fig 8) differed significant-ly from that of the number of combs pres-ented above: χ2calc = 8.12 > χ2tab 0.05 =5.99, P = 0.018 (df = 2), as well as fromthe number of brood cells: χ2calc = 6.76 >

χ2tab 0.05 = 5.99, P = 0.035 (df = 2). Thusthe number of foragers was not correlatedeither with the number of brood combs orwith the number of brood cells in these col-onies.

Significant differences were found in theflight activity of different colonies (table II).Colony No 6 was significantly more activein foraging than both Nos 3 and 5. Averag-

es for stinging during the same period (10-15 h) showed that bees in colony No 3stung significantly more than all 3 (June,23) or 2 other (Nos 4 and 5) colonies

(June, 24). Workers in colony No 4 stungsignificantly less than both Nos 3 and 6(June, 24). Thus the most active foragers(Nos 4 and 6) were not the most defensivebees, and the less active foragers (No 5)did not defend the colony less. The ratio ofthe number of stinging bees to the numberreturning to the hive varied considerably.According to the absolute number of sting-ing bees (table II; June, 24), no significantdifferences were found between coloniesNos 4 and 5 and between Nos 3 and 6;however, significant differences were

found between those colonies in the ratioof stinging bees. Averages for both daysare presented in figure 8. χ2-Test showedthat the number of stinging bees in col-onies Nos 5 and 6 did not differ significant-ly: χ2 = 0.73, P = 0.39 (df = 1). However,the ratio of stinging bees in these coloniesdiffered significantly: χ2 = 13.30, P = 0.00(df = 1). Statistical calculations did not show

a significant difference between averagenumber of stinging bees in colonies Nos 4and 5. However, the ratio of stinging beesto foragers in these colonies differed signif-icantly (fig 8). These results suggest thatthe ratio of the number of stinging to thenumber of foraging bees over the sameperiod of time could be used as a better in-dicator for a comparison of potential defen-sive behavior in different colonies than theabsolute number of stings left in a test ball.

Practical application

At the onset, examination of the bee col-onies was conducted with the examiner ful-

ly protected with bee veil and leather

gloves. The gloves had cuffs turned downwith the rough leather outside. This irritat-ed the bees considerably, and stimulatedthem to sting. Quite a lot of smoke wasused. Colony No 4 was examined secondafter another colony in this apiary waschecked. However, it defended itself so

vigorously that it was impossible to exam-ine all the combs. Some time later, when

gloves were not used and only a little

smoke was applied, it was possible to ex-amine that colony without any protectionand wearing only a short-sleeved shirt. Af-ter the results of this investigation showedthat smoke irritated the bees, the authorexamined the col-ony without using anysmoke, without any protection and withouta shirt. Twenty combs were taken out ofthe hive and were lifted one by one to faceheight. The examiner was only stung oncein the thorax during this operation. Manycolonies were subsequently examined in

the same manner. In the presence of the

beekeepers, the author opened hives andremoved combs without using protection orsmoke at different apiary sites in Ghana,such as Tamale in the north, Nkoronza,Sampa and Boabeng-Fiema in the centralregion and Accra and Winneba in the

south. When more laborious manipulationsare involved, eg finding a queen, the useof a little smoke in recommended. Howev-

er, the smoke should not be directed insidethe colony, either through the entrance orbetween the top bars. It should be appliedoutside to keep the bees inside the hives.The management technique for African

bees should differ from that for Europeanbees.

DISCUSSION

The purpose of this investigation was notto accurately characterize the defensivebehavior of the A m adansonii population,or the seasonal variation in colony de-fense. The aim was to determine whetherthe practical observations had some scien-tific confirmation.

The defensive behavior of a bee colonythroughout an entire day has not been in-vestigated previously. It was not common

knowledge that colony defense alters dur-ing the course of the day. Consequently,an account of the diurnal pattern of de-fense behavior has not been presented tillnow. In published papers, the hours of in-vestigation have generally not been includ-ed, or conclusions have been drawnbased upon comparisons conducted at dif-ferent hours.

The results are too limited to character-ize defensive behavior throughout bothseasons. However, the results clearlyshow that the defensive behavior of thesame colony did not remain constant, anddiffered significantly in different seasons.The pattern of diurnal colony defense, aswell as the ratio of stinging bees to forag-ers, did not differ within the same season;however, it differed significantly betweenseasons. Certain facts suggest that theconclusions should not be limited only tothe days of observation. Colony defenseon both days during the honey flow sea-son was similar, although observationswere conducted at an interval of about 1

month. Observations on flight activity con-ducted by Woyke (1992) in Ghana over a10-day period showed 2 distinct patternsof flight activity, one characteristic for thedearth period and the other for the honeyflow season. Because a high correlationwas found between flight activity and de-fensive behavior, it is suggested that 2 pat-

terns of colony defense, characteristic foreach season, may exist. The results pre-sented here agree well with the generalobservation that bees defend the coloniesmuch more vigorously during the honeyflow season than during the dearth period- summer rainy season.

To compare defensive behavior of col-onies of unequal strength, Collins and Ku-basek (1982) equalized them so that eachconsisted of 900 g of bees. Villa (1988)equalized the colonies several times by re-moving sealed brood from the strongestcolonies. Although these measures wereof interest, they involved much trouble anddid not assure the same number of flyingbees. Another possibility has been pres-ented here with the calculation of the ratiobetween foragers and stinging bees. Thecorrelation between colony defense andflight activity was not calculated until

Woyke presented his preliminary findings(1989) The 4 steps of colony defense pro-posed by Collins et al (1980) could not beinvestigated during this study, because theworker bees attacked the leather ball im-

mediately after it had been suspended in

front of the hive entrance.

CONCLUSIONS

A wide variation was observed in diurnaldefensive behavior. A typical pattern ofdiurnal defensive behavior was found toexist. A high positive correlation existed

between flight activity and defensive be-havior. The bees stung less during the pe-riod of low flight activity. The pattern of de-fensive behavior was very different duringthe non-productive and the honey flow

season respectively. The defense behaviorwithin each season was similar; however,it differed between seasons. During thehoney flow-winter dry season, the bees

stung about 4 times more than during thenon-productive-summer rainy season.

Bees in front of the entrance to the hivedefended the colony more vigorously thanthose that flew out from the combs. Thesmoke irritated the bees, resulting in moreactive defensive behavior. Significant dif-ferences existed in defensive behavior be-tween different bee colonies. The ratio ofthe number of stinging bees to that of theflying bees appears to be a better indicatorfor comparison of potential defensive be-havior of different bee colonies than theabsolute number of stings. It is recom-

mended that a comparison be made of de-fensive behavior of different bee coloniesat the same period of time. For practicalbeekeepers, it is recommended that workwith Apis mellifera adansonii be carried outduring periods of low flight activity.

Résumé — Variation journalière et sai-sonnière du comportement de défensede l’abeille africaine Apis melliferaadansonii au Ghana. Nous avons étudiéle comportement de défense de la coloniepar l’abeille africaine, Apis mellifera adan-sonii, afin de formuler des recommanda-tions pour les manipuler (quand, comment)en évitant d’être piqué. Une balle en cuirde couleur sombre a été suspendue pen-dant 1 min 50 devant l’entrée ou à 25 cmau-dessus des barrettes de 2 passagesouverts entre les cadres. On a compté lenombre de piqûres faites dans la balle,ainsi que le nombre d’ouvrières rentrant àla ruche durant 5 min. Les 2 comptagesont été effectués chaque heure, 13 fois parjour. Les résultats montrent une forte varia-tion du comportement défensif au cours dela journée. Il existe un schéma typique ducomportement défensif diurne. Pendant lasaison humide (été, période de disette), onobserve un pic le matin (fig 1) et pendantla saison sèche (hiver, période de miellée),un pic tôt le matin et un autre tard le soir

(fig 3). Il existe une corrélation fortement

positive entre l’activité de vol et le compor-tement défensif (figs 1-3). Les abeilles pi-quent moins lorsque l’activité de vol est fai-ble. Les abeilles devant l’entrée défendent

plus la colonie que celles qui quittent lesrayons (fig 3). Durant la saison sèche, lesabeilles piquent en moyenne 4 fois plusque pendant la saison humide (fig 6). Lecomportement de défense diurne varied’une colonie à l’autre (fig 7 et tableau I).Le pourcentage d’abeilles qui piquent parrapport aux abeilles qui volent est un

meilleur indicateur pour comparer le com-

portement défensif potentiel de différentescolonies que le nombre absolu de piqûres(fig 8 et tableau II). On recommande demanipuler les colonies d’A m adansonii

lorsque l’activité de vol est faible.

Apis mellifera adansonii / comporte-ment défensif / rythme d’activité / activi-té de vol / Ghana

Zusammenfassung — Tages- und Sai-sonvariation beim Verteidigungsverhal-ten afrikanischer Bienen (Apis melliferaadansonii) in Ghana. Das Verteidigungs-verhalten der afrikanischen Biene A melli-fera adansonii wurde untersucht, um Emp-fehlungen zu entwickeln, wann und wiediese Bienen ohne die Gefahr übermäßi-

ger Stiche bearbeitet werden können.Dunkle Lederbälle wurden 50 cm oberhalbdes Fluglochs oder 25 cm oberhalb vonzwei geöffneten Wabengassen für 1 min

aufgehängt. Die Zahl der im Ball zurückge-lassenen Stacheln wurde gezählt. Ebensowurde die Zahl der während einer Zeit-

spanne von 5 min zum Volk zurückkehren-den Trachtbienen aufgezeichnet. Beide

Zählungen wurden stündlich durchgeführt,13 mal den ganzen Tag über. Die erzieltenResultate zeigten eine große Variabilitätdes Verteidigungsverhaltens im Tagesver-lauf. Es bestehen typische Muster im tägli-

chen Verlauf des Verteidigungsverhaltens.Während der trachtlosen Zeit zur sommer-lichen Regenperiode wurde ein einzigerGipfel erhöhter Verteidigungsreaktion amMorgen beobachtet (Abb 1); zur Trachtzeitwährend der winterlichen Trockenperiodetraten zwei Gipfel auf, einer zeitig amMorgen und der andere spät am Abend(Abb 3). Es besteht eine hohe positiveKorrelation zwischen Flugaktivität und Ver-teidigungsbereitschaft (Abb 1, 2, 3). Wäh-rend schwacher Flugaktivität stechen dieBienen weniger. Die Bienen vor dem Flug-loch sind eher zur Verteidigung des Volkesgeneigt als Bienen, die aus den Waben-gassen des geöffneten Volkes auffliegen(Abb 3). Während der Tracht in der winter-lichen Trockenzeit stechen die Bienenetwa vier mal mehr als während der Som-

merregenzeit (Abb 6). Es besteht eine Va-riabilität im täglichen Verteidigungsverhal-ten zwischen verschiedenen Völkern (Abb7, Tabelle I). Der Prozentsatz stechenderBienen im Verhältnis zu den Flugbienen istein besseres Maß des Verteidigungspo-tentials eines Volkes als die absolute Zahlder Stiche (Abb 8, Tabelle II). Es wird em-fohlen, mit Völkern von A m adansonii zurZeit geringer Flugaktivität zu arbeiten.

Apis mellifera adansonii / Verteidi-

gungsverhalten / Aktivitätrhythmus /

Flugaktivität / Ghana

REFERENCES

Brandenburgo MM, Gonçalves LS, Kerr WE

(1977) Effects of Brazilian climatic conditionsupon aggressiveness of Africanized coloniesof honeybees. Cienc Cult (Sao Paulo) 29,255-280

Collins AM, Kubasek KO (1982) Field test of

honey bee (Hymenoptera: Apidae) colony de-fensive behaviour. Ann Entomol Soc Am 75,383-387

Collins AM, Rinderer TE, Tucker KW, SylvesterHA, Lackett JJ (1980) A model of honeybeedefensive behaviour. J Apic Res 19, 224-231

Collins AM, Rinderer TE, Harbo JR, Bolten AB(1982) Colony defense by Africanized andEuropean honey bees. Science 218, 72-74

Free JB (1961) The stimuli releasing the sting-ing response of honeybees. Anim Behav 9,193-196

Stort AC (1970) Methodologia para o estudo dagenetica da aggressividade de Apis mellifera.Congr Bras Apic 36-51

Stort AC (1974) Genetic study of the aggressive-ness of two subspecies of Apis mellifera inBrazil. 1. Some tests to measure aggressive-ness. J Apic Res 13, 32-38

Stort AC (1975) Genetic study of aggressive-ness of two subspecies of Apis mellifera inBrazil. 5. Number of stings in the leather ball.J Kans Entomol Soc 48, 381-387

Villa JD (1988) Defensive behavior of African-ized and European honeybees at two eleva-tions in Columbia. J Apic Res 27, 141-145

Woyke J (1973) Experiences with Apis melliferaadansonii in Brazil and Poland. Apiacta 8,115-116

Woyke J (1989a) Biology and management ofAfrican honeybees Apis mellifera adansoniiin Africa. XXXII Int Apic Congr, Rio de Janei-ro, 44-47

Woyke J (1989b) Experiences with queen beesrearing and instrumental insemination of Apismellifera adansonii in Ghana. XXXII Int ApicCongr, Rio de Janeiro, 61

Woyke J (1992) Diurnal flight activity of Africanbees Apis mellifera adansonii in different sea-sons and zones of Ghana. Apidologie 23,107-117