Apidologie 36 (2005) 169–185 © INRA/DIB-AGIB/ EDP Sciences, 2005 169 DOI: 10.1051/apido:2005014 Review article

The males of and other stingless , and their mothers1

Hayo H.W. VELTHUISa*, Dirk KOEDAMb, Vera L. IMPERATRIZ-FONSECAb

a Klemit 1, 5325 KG Wellseind, The Netherlands b Laboratório de Abelhas, Depto. de Ecologia, Instituto de Biociências, USP, Rua do Matão, Trav. 14, No. 321, CEP 05508-900, São Paulo,

Received 4 November 2004 – Revised 7 January 2005 – Accepted 17 January 2005 Published online 1 June 2005

Abstract – Female behaviour in social and the queen-worker conflict with respect to male production have been the focus of many studies. Although male production is an investment that is in conflict with investment in colony size, males play a vital role in colony reproduction. This paper reviews the production patterns of male stingless bees, their activities once they have reached adulthood and their origin (i.e., are they sons of workers or of queens). The existence of a broad spectrum of -specific patterns of male production, sex ratios, and male parentage offers ample opportunities to discuss the influence of ecology on the dynamics of stingless colony life. The paper also argues that selfishness causes the queen and the workers to compete and each to adopt certain strategies in their effort to produce male progeny. It is this competition, expressed in various forms during the characteristic and socially complex process of cell provisioning and oviposition, that could help explain the variable outcomes of male parentage at the species level as we currently know them.

male / sex ratio / life history / male aggregation / queen-worker conflict / provisioning and oviposition process

1. INTRODUCTION monopolize male production. This characteris- tic differs greatly between Melipona species. Stingless bees form an ancient (Michener This , therefore, offers ample opportuni- and Grimaldi, 1988) and rather diversified ties for studies on the factors that have had an (Michener, 1974, 2000) group of mass-provi- impact on the evolutionary rules governing the sioning eusocial bees. They vary considerably dynamics of colony life. This interaction in several of the characters for which sociobio- between our concepts of primary evolutionary logical theory would predict a basic uniformity. rules and the ecology of bees will be reviewed Trivers and Hare (1976) made clear that work- in this paper. ers of hymenopteran colonies, headed by a sin- Colonies of stingless bees are made up of gle monandrous queen, have reproductive males and females, and the latter are divided interests that are different from their mother. into workers and queens. Differences between This discord has its expression in the origin of males and females start with the fertilization of the males, some of which are sons of the queen the egg: the unfertilized egg becomes a male, whereas others are sons of some of the workers. while the fertilized egg is female. Sex determi- The workers of such colonies find their genes nation, therefore, is genetic and is related to better represented in sons and nephews than in . In contrast, the main decisive brothers and should, accordingly, attempt to factor in the development of a fertilized egg

* Corresponding author: [email protected] 1 Manuscript editor: Gudrun Koeniger

Article published by EDP Sciences and available at http://www.edpsciences.org/apido or http://dx.doi.org/10.1051/apido:2005014 170 H.H.W. Velthuis et al. into a queen or a worker is the quality/quantity colony characteristics. However, major differ- of food in the brood cell. ences exist in the way in which males are pro- Queens can be reared in different ways. duced. Most genera of the tribe Meliponini construct occasionally a larger brood cell that contains more food than the common brood cells. This 2. THE PATTERNS OF MALE quantitative factor modifies the differentiation PRODUCTION of the female in the cell: she will become a queen. The smaller and more common cells In general, male production in a social harbour workers and males. Thus, like in the colony is influenced by outside factors related honeybees, queens of these genera emerge to climatic periodicity, and factors inside the from specific queen cells. Different mecha- colony such as colony strength and demo- nisms are seen in the remaining genera. In Frie- graphic composition. Under temperate condi- seomelitta, for example, large queen cocoons tions, climate has a preponderant impact. The occur, but are the result of larvae perforating seasonal factors force colonies to produce their the wall of an adjacent cell. These larvae then sexuals during a brief period, thereby synchro- gain a second portion of food, enabling the nizing the colonies of a population. In contrast, modification of their development into a queen bees in tropical regions may be largely inde- (Terada, 1974; Faustino et al., 2002). All brood pendent of such climatic influences, although cells in the genus Melipona are approximately the alternation of dry and wet seasons may the same size. In this genus it is the minute var- cause males to be more frequent in one period iation in the quantity and, perhaps, also in the of the year than in the other. Synchronous pro- quality of the food that contributes to the devel- duction at the population level, therefore, may opmental differentiation between workers and become less pronounced and the role of within- queens. A two-locus genetic mechanism may colony factors more evident. also prevent 75% of the Melipona females from Male production is an investment that is in becoming queens (Kerr, 1950; Velthuis and conflict with investment in colony size: instead Sommeijer, 1991). Because the differences in of a worker a male emerges that does not par- the food for worker or queen development are ticipate in regular colony activities like cell so small, young Melipona queens emerge quite building, cell provisioning, colony defense, regularly even though hardly any are actually and foraging. He does, however, represent the needed. These superfluous queens are then colony in the reproductive arena. Investments killed by the workers. in reproduction may occur once the existence This peculiar and seemingly overproduction and future of the colony is assured. of queens and their fascinating mechanism for How would male production be regulated at caste differentiation have stimulated many the colony level? For a perennial colony with studies on Melipona (Kerr, 1950; Velthuis a long life expectancy, which characterizes and Sommeijer, 1991; Koedam et al., 1995; Melipona species, there may be two ways: (1) Wenseleers and Ratnieks, 2004; Wenseleers constantly as a proportion of the cells produced, et al., 2004). As a result and because a in a ratio that increases both with colony size number of Melipona species have been domes- and with the rate of cell production or (2) as an ticated, the genus Melipona is better known outburst of limited duration once the proper compared to most of the other stingless bee colony size and conditions have been reached. genera. In the first instance, male production is a con- Males are the prime subject of this study. We stant but light impediment. In the second case, will discuss the patterns of their production, male production may lead to distinct fluctua- their activities once they reach adulthood, and tions in the worker force, which forces the col- their origin (i.e., are they the sons of workers ony to recover after a male producing period. or of queens?). Alhough this paper concen- Long-term ecological factors, such as the prob- trates on the genus Melipona, when appropriate ability of the presence of suitable but unoccu- other genera will be mentioned. Melipona is a pied nesting sites in the habitat, may also be neotropical genus comprising about 40 species, involved. Such habitat characteristics deter- which, to a large extent, are uniform in their mine the incidence of swarming. If swarming Males of stingless bees and their mothers 171 is frequent, we might expect colony growth rate 2.1. Temporal patterns of male to be important and thus male production to be production absent or limited in smaller colonies. One pos- sible example of this is M. mandacaia, a species It has been documented at colony level of a that lives in the caatingas of the state of Bahia, number of species, that most of the males are Brazil. These areas are exposed to extremely produced in periods with a restricted duration. long periods of drought, during which bee col- We have termed this the Male-Producing Peri- ods or MPP (this term is preferred over the onies may undergo considerable size reduction. almost equivalent term Male Emerging Period Populations, therefore, are characterized by or MEP proposed by Chinh et al., 2003). In high colony-mortality rates. Once a rainy recent studies, this phenomenon has been doc- period has caused the vegetation to bloom, the umented to occur in M. asilvai, M. bicolor, M. colonies increase rapidly in size and then favosa, and M. subnitida as well as in Plebeia swarm to occupy the nesting sites that became remota, postica, Schwarziana vacant (Castro, pers. com.). We might expect quadripunctata and (Lepidotrigona) male production in such a species to be concen- ventralis (Bego, 1982; Koedam et al., 1999; trated in time at both the population level and Velthuis et al., 2002; Sommeijer et al., 2003; the colony level, because the impact of the Alves, 2004; Alves et al., 2004; Chinh, 2004; environment is so great that the within-colony Alves, unpubl.). It is, therefore, a wide-spread factors have little influence. In contrast, there event. MPPs may be the result of a synchro- might not be a vacant nest cavity for long peri- nous, albeit differential, production of repro- ods of time for species living in areas where the ductive eggs by a number of workers, some- environment is more stable. Since most times with the production of haploid eggs by Melipona species live in more stable environ- the queen, over a short period of time (days or ments and build their nests inside living trees, weeks) (Koedam and Imperatriz-Fonseca, unoccupied cavities of the proper size might be 2004a; Koedam et al., 2005). rare. These colonies may, therefore, have to Even though individual colonies may pro- delay reproduction through swarming. As a duce their males in MPPs, these periods may result, they may only reproduce through their not be synchronous among colonies and thus, males, one of which may find a young queen at the population level, male production is usu- that supersedes an old one in another colony. In ally not restricted to certain periods of the year. such a case, small colonies may also produce This was shown in studies on M. favosa (Chinh males as their continuous existence is not at et al., 2003) (Fig. 1) and M. bicolor (Alves, stake. For the genus as a whole, swarming is 2004). MPPs within individual colonies, there- characteristically infrequent (Roubik, 1989). fore, have no impact on male presence at the We consider this ecological factor of great population level and do not affect queens that importance for Melipona biology. are on mating flights. It also appears that MPPs

Figure 1. Male production periods in six colonies of Melipona favosa in Tobago. Solid bars represent the peri- ods of male emergence in these colonies. Reproduced from Chinh et al., 2003. 172 H.H.W. Velthuis et al.

Table I. The investment in sexuals as the proportion July, these values were 11.3% for males and 9% of the total brood of some well-studied species. for queens. Queens were present in all combs, species percentage authors usually in low percentages. Males were often absent or occurred in rather low numbers. If Melipona asilvai 26 Alves, unpublished they were more frequent, their percentage usu- M. beecheii 37.5 Moo-Valle et al., 2001 ally remained below 30%; the maximum found was 36%. This indicates that male production M. bicolor 12.5 Alves, 2004 is clustered in time. This result confirms the M. favosa 22.5 Sommeijer et al., 2003 more detailed observations of Koedam et al. M. subnitida 15 Alves, unpublished (1999), who found only 31 males on 14 combs having 2 496 cells. Of these 31 males, 29 were M. trinitatis 13 Sommeijer et al., 2003 produced in only a few combs. 29 Alves, unpublished To further study male production, Koedam Schwarziana 20.5 Alves, unpublished et al. (2005) manipulated food circulation in quadripunctata two M. subnitida colonies. Before manipula- tion, these colonies produced 3% and 0% males and 9.2% and 8% queens. During the manipu- of different species have different intensities, in lation period, the male proportions increased to number as well as duration. 47% and 25% respectively, while the queens When there is an MPP during which differ- scored 9% and 18% resp. These results show ent individuals lay reproductive eggs, it is not that queen numbers remained rather stable necessary that a mechanism be present that ena- despite manipulation, while male numbers bles the workers to recognize whether the fluctuated. In other words, male production queen's egg has been fertilized or not (Nonacs was subject to colony conditions. and Carlin, 1990) and to which they respond by Ten M. beecheii colonies were studied for laying unfertilized eggs. A general condition or one year by Moo-Valle et al. (2001). The pop- signal inside the colony related to the start of ulation produced brood that was 22.9% male, an MPP would suffice to stimulate both work- and 14.6% of the female brood were queens. ers and queens to lay haploid eggs. The average male production by the entire pop- ulation was constant, except for September when it was only 11%. There was, however, a 2.2. MPPs and the ratios of males versus large monthly variation in the frequency of workers and queens males between the colonies, much larger than Some population-wide studies on the pro- between seasons. The percentage of queens duction numbers of males, workers, and queens fluctuated over the year between 11% and 21%, have been conducted that cover a long time being low in July and high in May. In a factor span. They included the following species: M. analysis, the factor ‘season’ was of equal quadrifasciata (Bezerra, 1995), M. subnitida importance as the factor ‘colony’ in explaining (Koedam et al., 1999, 2005; Alves, unpub- the fluctuation. The average queen/male (q/m) lished), M. beecheii (Moo-Valle et al., 2001), ratio was 1/2. A separate experiment with six M. favosa (Koedam, 1999; Sommeijer et al., colonies studied the effect of food manipula- 2003; Chinh, 2004), M. trinitatis (Sommeijer tion. The three control colonies were fed during et al., 2003), M. asilvai (Alves, unpublished), the experiment, while the three experimental and M. bicolor (Alves, 2004). colonies were not. Food stored in the experi- mental colonies was reduced to one-third that The results of those studies are summarized in the control group. After a year, these exper- in Table I. imental colonies were only one-third the size of Alves analyzed 29 M. subnitida combs from the control colonies. The latter were identical different colonies, harvested in April, June and to the sample of 10 colonies. The brood of the July 2003 in their natural habitat. In April, experimental colonies consisted of only 1% males constituted only 2.4% of the brood and drones, while queens made up 10% of the queens 10% of the females. In June, the average female brood, thus the q/m ratio was 1/0.1. In male score was 7.5% and that of queens 6%. In comparison to natural conditions, food limitation Males of stingless bees and their mothers 173 affected the proportion of drones much more ated between 4.3% and 9.2% per month. Males than the proportion of queens, probably as a constituted a monthly average of 3.7% (1.7– consequence of colony size reduction. 6.7%, 10 months) with the exception of August In M. favosa (Sommeijer et al., 2003), 167 (23.9%) and November (12.5%). At the colony brood samples, taken from 78 colonies over a level, male frequencies were found to fluctuate number of years, contained 16 342 pupae. Of much more than queen frequencies and to differ these, 2 827 were males (17.3%), 689 were much more between colonies sampled at the queens (4.2%), and 12 826 were workers same time. More than 50% of the males were (78.5%). Due to the occurrence of male-pro- aggregated on a single comb of their respective duction periods, the percentage of males in a colony, making up 24–92% of all the cells of given comb could be as high as 74.2%, but was that comb, while 70% of the combs contained often 0% (in 26.5% of the samples). On an less than 5 males each. annual basis, queen production varied less (4– Comparing the reproductive investments of 6%) with exceptions in March (8%) and July these species, the percentages of males found (7%). Another study of the same species varied from 6% to 23%, those of queens from (Chinh, 2004) included 6 colonies at the same 5% to 15%. Taking both sexes together, M. location, studied from October to February. bicolor and M. trinitatis were the most prudent These colonies produced 639 males during a (Tab. I). They invested 13% of all brood in period of 119 days. Males were concentrated in reproduction. In contrast, M. beecheii was the some of the combs (Fig. 1) and constituted other extreme with 37.5% of all brood being about 8% of the total brood production, con- sexuals. Apparently, there are MPPs for all of trasting with the annual average of 17.3% of the these species, and the colonies of a single spe- Sommeijer et al. study. The author explained cies are not synchronized in their male produc- this difference by pointing out that, according tion. Also, the maximum proportion of males to Sommeijer et al. (2003), males were most in the brood sample varies among the species frequently encountered in the months July and and, consequently, the fluctuations in worker August, reaching populational averages of 25– numbers are more prominent in some species 30%. than in others. Similar figures exist for M. trinitatis Queen/worker proportions appear to be (Sommeijer et al., 2003). This species was more uniform. In the abovementioned studies, sampled for 4 months: 21 combs from 15 col- a synchrony in the quantitative aspects of the onies contained 2 567 pupae. Queens were production of the two sexual forms was not present in 80% of the samples, males in only reported. However, Koedam (1999) did find a 45%. Of the females 5.2% were queens. Males positive significant correlation for M. favosa, amounted to 7.7% of all brood, but made up between the numbers of queens and males per comb. This indicates that a common environ- 31% of the brood of a single comb. mental factor may be involved in the produc- Alves (unpublished) counted the cells in tion of queens and males. 15 combs of M. asilvai, collected between 15 April and 19 July 2003. She found a total of 1 719 cells: 302 males, 102 queens and 2.3. What regulates the occurrence 1 315 workers. Except for 3 combs, the average of an MPP? proportion of queens in the female brood was What kind of information is available to a 5.8% (4–8%). When all 15 combs were member of a colony that determines when included, the proportion was 8.4%: one small males can be produced without much harm to comb contained 23% queens. Males, amount- the future performance of the colony? We pro- ing to an average of 17.6%, were either absent pose that demographic factors play a role (see or numbered less than 5 in 8 of the 15 combs. also Bego, 1982). Tasks like constructing and A sampling of 18 M. bicolor colonies provisioning brood cells are performed by (Alves, 2004) during a year period concerned younger workers, while eggs are produced by a total of 97 combs containing 11 745 brood the queen. We suggest that there is a certain cells. Queens made up, on average, 6.5% of the adjustment in the production capacities of each female brood. At population level, this fluctu- of the parties, i.e. for the numbers of cells and 174 H.H.W. Velthuis et al. eggs produced, and that it is conceivable that at being produced and both queen and workers a certain rate, the cell construction capacity is may respond to favourable colony conditions, higher than the egg production capacity of the the queen and her workers are competitors in queen. In these circumstances, there will be a the realization of this number of males. The surplus of young workers. By laying haploid outcome of this competition and the strategies eggs in part of the cells that the workers pre- applied vary greatly between the species. For pare, the number of workers is down-regulated instance, all M. beecheii males appear to be the for the near future, possibly reaching an equi- sons of the queen (Paxton et al., 2001), while librium with the egg-laying capacity of the 95% of the M. favosa males are sons of the queen. Whether this is a fine-tuned process, workers (Sommeijer et al., 1999). Tóth et al. keeping worker numbers stable, that can be (2002) already discussed part of this plasticity governed by the queen alone or whether it in male production. causes notable fluctuations in worker numbers, All of the above concerns queenright condi- in which perhaps laying workers also take part, tions. In queenless conditions, the workers of is left undecided for the moment. a number of species produce eggs. This is out- Alternative solutions for repairing such a side the scope of the present study. disequilibrium in the production rates of cells vs. eggs include an earlier shift of the workers from nursing to foraging activities, which leads 3. THE VARIATION IN WORKER to an increase in food stored inside the nest, and OVARY FUNCTIONS IN THE swarming. The respective limiting factors of OTHER GENERA these alternatives are and availa- bility outside and storage space inside the nest, Annoted above, there is a variation in the and the presence of a new nest site for the proportion of males that are sons of workers. swarm. Since stingless bees first find the new Moreover, there are genera in which reproduc- nest site and then prepare for swarming, both tive workers are fully absent even though the alternative solutions depend on ecological fac- conditions concerning monogyny and monan- tors and may not be available for a given col- dry are met that, theoretically, would favour ony. worker-produced sons. However, this does not mean that the worker ovaries are disfunctional. 2.4. Fitness aspects of male production Apart from the fertile, reproductive eggs, If specific colony conditions favour the workers of many stingless bee species generate occurrence of an MPP and information on these trophic eggs. Trophic eggs are essentially infer- conditions is available to all colony members, tile (Akahira et al., 1970; Cruz-Landim and then the queen as well as some of the workers Cruz-Höfling, 1971; Koedam et al., 2001) and may respond by starting the MPP. At this are specialized vehicles for the transfer of pro- moment a divergence in the fitness interests of tein from a worker to the queen, comparable to queens and workers develops. A reproductive the glandular secretions that pass from a hon- worker shares 25% of her genes with the son eybee worker to the queen during trophallaxis of the queen, 37.5% with the son of a sister and (Velthuis, 1993). 50% with her own son. So, in conformity with In Geotrigona (Lacerda and Zucchi, 1999) Trivers and Hare (1976), we might expect the and in Tetragonisca (Koedam et al., 1996; workers to produce the males. Indeed, monan- Grosso et al., 2000) the workers are capable of dry is the rule in stingless bees (Peters et al., laying trophic but not reproductive eggs. In 1999) and in all Melipona species workers Trigona cilipes workers have trophic eggs in are capable of reproducing directly; being their ovaries, but do not lay them. In these gen- unmated, all of their eggs are haploid (Sakagami, era, therefore, workers may have active ova- 1982). Since the survival of the colony is in ries, but remain without offspring. Leurotri- their interest, workers should restrict their egg gona (Sakagami and Zucchi, 1974; Terada, laying in number and to time periods when col- 1974) and Friesella, on the other hand, have no ony survival is not at stake. Given that the col- trophic eggs, but do have reproductive eggs. ony can endure only a certain number of males Only Friesella, however, actually lay these Males of stingless bees and their mothers 175 eggs (Camillo-Atique, 1971, 1977; Imperatriz- The second explanation for the unusually Fonseca and Kleinert, 1998). In even more con- high frequency of young queens is that the inci- trast to the basic theory, the ovaries of Frieseo- dence of swarming is low, that mature colonies melitta (Zucchi, 1993), Duckeola (Sakagami are usually prepared for swarming already long and Zucchi, 1968) and Tetragonula (Suka and before they can actually do so, and that an Inoue, 1993) workers are never activated. In investment in queens is needed to maintain the fact, the ovaries of Frieseomelitta workers are preparedness of the colony. With this explana- degenerated (Boleli et al., 1999). The complex tion, we try to find the adaptive side of the situation of stingless bee worker ovaries and phenomenon at the colony level. Before a com- the eggs they may produce has been excellently parison with male investment is made, the reviewed by Silva-Matos et al. (2000). continued preparedness, in terms of queen numbers, should be added to the number of workers participating in the swarm. We can 4. SEX RATIOS expect, therefore, that on average, there should be only a small male bias in the q/m 4.1. Primary sex ratios ratio. In the other Meliponine genera, those with The ratio of queens/males at emergence var- specific larger queen cells, we find the more ies in the long-term brood samples of the usual situation for social . In a large sam- Melipona species of Table I: 10 for experimen- ple of Plebeia remota, Alves (unpublished) tal M. beecheii colonies, 1.1 for M. subnitida, found queens in 0.04% of the cells and males in 0.9 for M. bicolor, 0.6 for M. trinitatis, 0.5 for 25%; the primary sex ratio was 0.0016. In a natural M. beecheii, 0.3 for M. asilvai and 0.25 smaller sample from another population, the sex for M. favosa. This is interesting, because one ratio was 0.005. In more than 14 000 Schwarz- would expect the primary sex ratio to be iana quadripunctata cells, Alves (unpublished) strongly male-biased in a social insect that mul- found 2 900 males, 14 royal cells, and an addi- tiplies its colonies through swarming: total tional 67 miniature queens emerging from investments in males on the one hand and in the worker cells. Sometimes miniature queens entire swarm on the other should be in equilib- become physogastric and are able to head rium (Fisher, 1930). In these Melipona bees, it a colony like normal queens (Ribeiro and appears as if the workers accompanying the queen during swarming do not count. Alves, 2001). As the functional difference between these two types of queens remains There are two explanations for this unusual unknown, the primary sex ratio is 0.028 for all sex ratio. The first is that, theoretically, there is queens or 0.0005 for queens from royal cells an overproduction of queens, a maladaptation only. considering the colony level, that may be the consequence of self-determination (Ratnieks, In relation to the higher proportions of 2001). In classical terms, this is the interaction queens among the female brood in Melipona between a genotype capable of developmental and the proposed ecological explanation we plasticity and a variable environment, in cases have presented, it must be noted that there are where we do not see the impact of this environ- species in other genera that also have a low inci- mental variation. It is the expression of a con- dence of swarming. Scaptotrigona, for instance, flict between individual selection and selection lives in habitats similar to those of many at the colony level. Under such conditions, lar- Melipona, but appears to swarm more often than vae are expected to “decide” more often to Melipona. Virgin queens in a number of become a queen. As already mentioned, Kerr these genera are often maintained for longer (1950) proposed a genetic mechanism that time periods then in Melipona, sometimes renders 75% of the female larvae of Melipona “emprisoned” in special cells (Engels and species insensitive to the food stimulus that Imperatriz-Fonseca, 1990). This may be leads to queen development. The relationship another way for the colony to remain prepared between genome and food condition in the for swarming. Apparently, there is also more remaining 25% of the female larvae is variation in mating age for these queens then for unknown. those of Melipona. 176 H.H.W. Velthuis et al.

4.2. Operational sex ratios We can obtain a rough estimate of the mat- ing probability for males of the other species by The sex ratio at emergence is not very supposing that a new comb is started about important for the situation at mating. Mortality every 6 days. This means 60 combs a year. rates transform the ratio at emergence into the Using the data for the number of cells per comb operational sex ratio, i.e. the situation in which and male frequency (see Tab. I for refs.), we can a queen and a male will find each other and then calculate the male numbers per colony per which might shape their species-specific year to vary from 365 in M. subnitida to 1 000 behaviours. Almost no Melipona queens will in M. favosa. Depending on the incidence of survive after their emergence and have a swarming and the replacement of old queens, chance to partake in a mating flight (Imperatriz- the mating probability for a male might indeed Fonseca and Zucchi, 1995; Koedam et al., be in the order of 0.01–0.002. 1995; Wenseleers et al., 2004). As can be con- cluded from the long residency of marked queens, laying queens of Melipona may remain 5. LIFE HISTORY OF MALES in their colony for over a year. Substitution is infrequent. The incidence of swarming is also Considering the very low probability that a low, far less than once a year in most species. male will ever mate, mating does not appear in Taking substitution and swarming together, a activity-time budgets of males. In fact, mating queen will depart from a colony for a mating has hardly ever been seen. Males stay inside the flight on average only once every 8–12 months. colony until the age of about 2–3 weeks (P. How many males will be available for a droryana, Cortopassi-Laurino, 1979; M. beecheii queen to select a mate from? Chinh's study and M. favosa, van Veen et al., 1997). During (2004) indicated a productivity of 639 males that time, they have been reported to be over a period of 119 days, originating from involved in building activities, trophallaxis, 6 colonies, i.e., on average, 0.895 males per nectar dehydration (Imperatriz-Fonseca, 1973; colony per day. If the period of sexual activity Cortopassi-Laurino, 1979), incubation, and nest of males lasts 14 days, each colony in the area defense (Kerr, 1990). According to Nogueira- has, on average, 12.5 males on the wing. The Neto (1997), some of these activities (building, operational sex ratio largely depends on the nest defense) are only incidental, even though number of colonies in the area. If there are 10, males have been observed to secrete wax. Nec- for example, the queen may find 125 males tar dehydration could also play a role in their from which to choose. A few other queens preparation for leaving the nest and thus not be searching for a mate at the same time would a social activity. Van Veen et al. (1997) found hardly make a difference. Since all species are that males of M. beecheii spent more than 90% monandrous, each of the queens reduces the of their time walking, standing, self-grooming, number of males by just one. Thus, plenty of and stirring. males remain, provided the males that spend Once the males have left the colony, not their energy in a fruitless pursuit of a queen much is known of their activities. They do not soon recover and again pursue with similar vig- return inside the colony, so it is supposed they our, if given a second chance. live the life of a solitary male bee. Their life For males, however, this operational sex expectancy is unknown, but, based on survival ratio is too optimistic a figure, because many under laboratory conditions, they might obtain of them are mature in time periods when there an age of 4–6 weeks. is no virgin queen available. Again based on While they live on their own, males have to Chinh’s study, the annual production of a M. forage and may visit the same flowers as work- favosa colony is around 320 males. With a sup- ers of their species. Roubik (1990) suggested posed annual production of just a single queen that males may follow workers back to their out for mating, the probability for a male to nest and thus learn the locations of conspecific mate is only 0.003. This is an optimistic per- nests. There have been made several observa- spective, given that the estimate of male pro- tions of stingless bees males that were lured to duction based on Chinh (2004) is well below orchids, where they do not find food but that of Sommeijer et al. (2003)! become loaded with the pollinia of the orchid Males of stingless bees and their mothers 177 during an act of pseudocopulation (Singer (Paxton et al., 2003; Imperatriz-Fonseca et al., et al., 2004). There is little doubt that males do 1998). Studies by Paxton (2000) and Cameron play a role in the of these orchids as et al. (2004) have shown that the congregating well as other flowers. males come from a large number of colonies. Inbreeding, therefore, is largely prevented. Sommeijer and de Bruijn (1995) described a 6. HOW DO MALES MEET male congregation of M. favosa at some dis- A QUEEN? tance from a nest. This group of males had gath- ered on a wall, a site that was also visited by the Theoretically, all of the mating tactics known virgin queens. The authors reported that the for other bees (see Thornhill and Alcock, 1983; males and virgin queens flew off together. Mat- Paxton, 2005) can be applied to stingless bees. ing could not, however, be observed. T. angus- Are there species, for example, where males tula (Nogueira-Ferreira and Soares, 1998) and maintain a territory or where they jointly attract P. remota (Alves et al., in prep.) males were females to a lek? We do not know. There is, reported to aggregate at a new nesting site a few however, a kind of mating strategy that is quite days before a young queen arrived with a conspicuous and therefore observed for a swarm. The males stayed there until the nuptial number of species: the temporary congregation flight had taken place. of males near a colony. Such male congrega- Long-lasting nuptial flights are known for tions were described already long ago. More M. quadrifasciata (Silva et al., 1972) and M. recently, descriptions have been made for Ple- beecheii (van Veen and Sommeijer, 2000). In beia droryana (Cortopassi-Laurino, 1979), both cases, the queens were observed to leave (Engels and Engels, the colony in rapid flight and to return, some- 1984; Engels and Imperatriz-Fonseca, 1990; times with a mating sign, up to 100 and 40 min. Paxton, 2000), Trigona () dorsalis later, respectively. Such a long duration of a (Roubik, 1990), Tetragonisca angustula mating flight indicates that queens also search (Imperatriz-Fonseca et al., 1998; Nogueira- for males instead of encountering them at the Ferreira and Soares, 1998; van Veen and nest entrance. Sommeijer, 2000), Trigona collina (Cameron et al., 2004), M. fasciata (van Veen et al., 1997), and were seen in M. crinita, M. scutellaris, M. 7. WHO PRODUCES THE MALES? subnitida, Nannotrigona testaceicornis, Trigona spinipes, Paratrigona subnuda (Cortopassi- Recently, male parentage in stingless bees Laurino, unpublished). Males of different spe- was reviewed extensively (Tóth et al., 2004). cies may even be found together. In all of these We suggest that the reader should refer to that cases, it was often noted that the males clustered paper for most of the literature on the subject. close to a colony with a queen to be inseminated It shows that worker reproduction is far more or nearby a queenless nucleus that just a few common in the genus Melipona than in the days earlier was separated from the mother majority of the other genera. At present, there colony. Apparently, these colonies produce is no satisfying explanation for this higher fre- an extremely attractive odour. In fact, 30– quency or for the important differences 50, sometimes even hundreds, of males have between the species of Melipona. The Tóth been counted near such colonies, supposedly et al. paper discusses 10 predictions for worker arriving from great distances. They cluster reproduction. Since this subject is of great together (Scaptotrigona) or remain more dis- interest from the theoretical point of view of persed (Tetragonisca) near the nest entrance, a sociobiology, these predictions deserve our smaller number may even fly around the comments. Some of the arguments are rather entrance, all awaiting the appearance of the abstract, while others are of a more classical queen. Such clustering begins just a few days biological nature. This leads to an overlap in the before the mating flight and disperses a few days predictions. In the monogynous colony with a thereafter. Once the virgin queen departs from monandrous queen, theory predicts that work- the nest, the males compete to mate. Some ers, having fitness interests different from those queens may even be mated by two males of the queen, should attempt to monopolize 178 H.H.W. Velthuis et al. male production. However, like in the dis- batches, and that MPPs occur because, in the cussed by Trivers and Hare (1976), a solution demography of the colony, a situation can arise is often found between the optima for workers in which the queen cannot keep up with the versus queens. One group of predictions con- number of cells workers are able to produce. centrates on cost/benefits, focusing on the indi- This leads to changes in the details of the pro- vidual egg layer as well as on the colony as a visioning and oviposition process (POP), such whole, because all colony members have their as a more rapid provisioning and a more inclusive fitness interests. The other set of pre- delayed egg laying by the queen and to a lower dictions concentrates on the instruments with degree of involvement of the workers of the which one party might control the other, i.e., the appropriate age class. These factors might have direct fitness interests. These instruments, consequences for the individuals involved. however, are also the means by which the cost/ While such a cause-and-effect chain could benefit ratio is controlled. It is through the joint explain the overall pattern of male production effects of these instruments that, in Meliponine within the genus, it is through the amplitudes bees like in the ants, a compromise is found of the process that we might find an explanation between the conflicting interests of the parties. for the differences between the species in the This compromise is not a static mean value, it relative amounts of males in an MPP. Such dif- is dynamic, leading to large variations around ferences include the average percentage of the mean. The mean values, therefore, may males as well as the proportions produced by vary over time and between species. either the queen or the workers. When meas- How can we achieve an understanding of the ured with molecular techniques, however, interactions of these instruments? Tóth and co- these kinds of data are still only the net result workers compared observational and molecu- of interactions. In fact, the “arms race” among lar genetic methods and preferred the latter the participants or the “power” of any of the because (a) the observer might confuse the parties could include the elimination of an trophic and the reproductive eggs of stingless opponent’s offspring. bees; (b) worker egg laying might be cryptic; It must be stressed that a different physiol- (c) behavioural observation is time-intensive, ogy and behaviour are needed to lay a trophic therefore observations are often restricted in or a reproductive worker egg. A Melipona time with the consequence that variation could worker needs five to six days to produce and lay escape detection; and (d) queens may also lay a trophic egg, while reproductive eggs may be male eggs. Genetic studies should circumvent laid with intervals of only one day (Koedam these problems, especially if a queen and a et al., 2001; Koedam, unpublished data). In worker lay an egg in the same cell and a male general, a worker laying a trophic egg does so emerges (who laid the egg?). The criticism of prior to egg laying by a queen. She then flees the behavioural method concerns the quality of from the cell. A worker that lays a reproductive the observations, not the method as such. In egg stays on the cell and closes it, resisting the fact, several molecular genetic studies, like queen or other laying workers that try to several behavioural observations, have also remove her (Sakagami, 1982). Morphologi- missed the variation or presented no explana- cally and physiologically, the two types of egg tion for the variation found. We believe that it also differ (see Koedam et al., 2001). The dis- is only through behavioural observations that tinction, however, does not imply that an indi- the variation, both within and among species, vidual should lay either trophic or reproductive can be interpreted. The understanding will eggs; some individuals lay both types in bring the cost/benefit ratios and the mecha- distinct phases of their life (Koedam and nisms involved together. In order to improve Imperatriz-Fonseca, 2004b; Koedam, unpub- understanding, the two methodologies are not lished data). of different ranks but are instead supplemen- A few examples of such intricate interac- tary. tions that might remain obscure in genetic stud- We have mentioned already that, in the ies may illustrate our case. They revolve around genus Melipona, great differences exist in the the POP, subdivided into a pre-provisioning proportion of males originating from worker- phase, a provisioning phase, and a post-provi- laid eggs, that males are often produced in sioning phase, followed by oviposition by the Males of stingless bees and their mothers 179 queen and, subsequently, cell closure by a sons. This suggests that direct competition is worker (see Zucchi et al., 1999, for definitions the driving force. Cells to oviposit in are scarce of POP elements). All of these phases are char- for Melipona workers. acterized by interactions among the parties. In It should be underlined here that the queen normal situations the queen will lay her egg responds to the egg laying of a worker, prior to after the cell has received enough food and a the moment she would do so herself, by vigor- worker will close the cell. Beig (1972) ously tapping the worker. Usually, this is not observed that this closing worker could be a successful. If the worker lays after the queen, reproductive one and that her egg is then in the however, the latter may notice but not respond; same cell as the one of the queen. He concluded often she passes by and may antennate the that 85% of the males in Scaptotrigona postica worker during egg laying and cell sealing as were sons of workers. Bego (1982) studied the well as during the re-opening of cells. We are same species and found that, during brief peri- not certain whether the queen notices that ods, up to 70% of the cells contained 2 or more worker reproduction occurs, but we do not eggs. Paxton et al. (2003), using genetic mark- exclude it. ers, concluded that almost all males were sons of the queen. Why does the queen allow a reproductive worker to lay an egg and to monopolize that Laying an egg while sealing the cell also cell, before she herself has oviposited? Expla- occurs in many species of Melipona. Competi- nations such as the queen is aging, tired, or tion among the hatched larvae determines what unattentive (Camillo-Atique, 1977; van Buren will eventually be produced. Another mecha- nism is that the reproductive worker lays her and Sommeijer, 1988; Sommeijer and van egg before the queen can do so, thus preventing Buren, 1992) are not appealing. Queens depend the queen from laying. This was observed in M. on trophic eggs for their nourishment, eggs that favosa, M. subnitida, Scaptotrigona barro- are laid by workers once the provisioning is coloradensis, and S. postica. This sequence completed. The Melipona queen may solicit prevents the two descendants from competing such egg laying by withdrawing from the cell. for food and implies that no eggs are lost. There Sometimes, she may have to wait for minutes is also a third possibility: namely that a repro- for a trophic egg to be laid. A reproductive ductive worker disturbs the cell closure of worker may then take advantage of the situa- another worker. This has been reported in M. tion and lay an egg before the queen can do so. bicolor, Friesella schrottkyi, and S. postica. A The worker then seals the cell. The queen, fourth mechanism of reproduction is that the being physically inferior (she has a smaller worker opens an already sealed cell some time head and thorax compared to workers and a after it is closed, occasionally eats the egg, ovi- much heavier abdomen), cannot prevent this. posits, and seals the cell again. These different There are a few ways in which the queen strategies are illustrated in Table II. may interfere with worker reproduction. One is With regard to the rates of cell production passive: i.e., she also produces haploid eggs and the rates of egg laying, it is only when a during an MPP. In M. subnitida, for instance, worker lays her egg prior to the queen that no plenty of cells still remain available for ovipo- eggs are lost. Indeed, the only study that com- sition by the queen despite the activity of repro- pares the rates before and during the MPP ductive workers. One study by Koedam et al. found that the egg-laying rate of the queen (2005) showed that two-thirds of the males remained constant while the cell construction were the sons of the queen (or 3:2, Contel and rate increased (Koedam et al., 2005). Kerr, 1976). Outside the MPP, only a few males Interestingly, it is always a reproductive are produced, all by the queen (Fig. 2, from worker that closes the cell, although theory Koedam et al., 2005). The same low number of would allow her to leave this task to her non- males produced outside the MPP has been reproductive sisters! This demonstrates that reported for other species. They are absent, there is selfishness involved: reproductive however, in M. favosa. workers competing for the use of the same cell, There appears to be an even more intriguing thereby destroying each other's eggs, diminish interference in M. bicolor. During an MPP the proportion of males that could be workers’ many reproductive workers may be present at 180 H.H.W. Velthuis et al. t colony Authors ying. See Zucchi et al. et Zucchi See ying. which to that the of queen laid next her egg tion process (POP) under queenrigh has been sealed Some time after cell e the cell immediately following on their egg la on their egg e the cell immediately following by another worker another by While cell is sealed While cell ing the provisioning and oviposi ing the provisioning ve workers occasionally excluded the queen. excluded occasionally workers ve queen oviposition Immediately afterImmediately a reproductive worker, thus containing a single egg, cells in and egg, a single containing thus worker, a reproductive ch trying to operculate after its oviposition. nononono yes yes yes no ? ? yes no ? ? yes (&) 100% 1977 da Silva, 2002 et al., Tóth and Kleinert, Imperatriz-Fonseca 1977 1998 da Silva, 100% no no no 1994 et al., Suka no ($)no 28.2% 30.8% 41.0% data. et unpubl. al., Koedam 66.7% 33.3% no no et in press al., Koedam yes (§)yes < 0.2% least at no (#)yes no yes2003 et al., Chinh yes yes1990 Bego, am et al. (2001) showed that reproductiam et al. (2001) showed as two days after cell provisioning. after cell days as two ous species. As a rule, such workers operculat ous species. As a rule, such workers excluding the queen excluding The laying of reproductive eggs by stingless bee workers dur bee workers stingless by eggs of reproductive The laying Speciesprovisioning, After cell Melipona postica S. sealed. was eggs, two having after which now the cell, worker, a reproductive M. favosa M. marginata M. subnitida M. quadrifasicata schrottkyii Friesella subnuda Paratrigona barrocoloradensis Scaptotrigona $ A former Koed study on this species by could occur as oviposition late & Worker by in and sealed oviposited cells exclusively This includes # § Various workers may compete for oviposition, ea may compete workers for oviposition, § Various (1999) for definitions of POP elements. (1999) for definitions conditions as documented for vari conditions Table II. Table Males of stingless bees and their mothers 181

25 Thus, the battle is fought almost without

20 overt aggressiveness between queen and worker. Only in M. subnitida was it once 15 observed that a worker pulled the queen from

10 a cell by grasping her mandibles. The worker then laid her own egg and sealed the cell. In 5

number of males produced Melipona, and also in a number other genera, 0 the outcome depends on the environmental 6 7 8 9 10111213141516171819202122232425 consecutive days (April, 1998) details within the colony as a whole, details

10 which change constantly. In this dynamic con- 9 text, each of the parties has successes and fail- 8 7 ures. If we had a nose as fine and sensitive as 6 bees, we would certainly be impressed by the 5 4 variety of signals that contain messages of 3 threat and submission. 2

number of males produced 1 0 15 16 17 18 19 20 21 22 23 24 25 26 27 ACKNOWLEDGEMENTS consecutive days (April, 1998) Our studies of stingless bees have received long- Figure 2. The sharing of the production of males by lasting financial support from FAPESP (São Paulo) the queen (open bars) and workers (closed bars) in and from the Dobberke Foundation (Amsterdam), two laboratory-kept colonies of Melipona for which we are grateful. While preparing this subnitida, over a period of up to two weeks in the paper we had stimulating discussions with, and month of April 1998, São Paulo. Up to the date obtained information included in this paper from when sharing began, in both colonies the queens Dr. Paulo Nogueira-Neto, Dr. Marilda Cortopassi- produced males occasionally. For details, see Laurino, and Denise de Araujo Alves. Furthermore, Koedam et al., 2005. we acknowledge the suggestions made by the refe- rees for the improvement of this manuscript. Lin- guistic help was given by Laura Cobb. a given cell at a given time. They repeatedly Résumé – Les mâles des abeilles Melipona, et des push each other from the cell, before as well as autres abeilles sans aiguillon, et leurs mères. Cet after worker oviposition. They also re-open article considère le modèle de reproduction et l’ori- cells in which a worker has oviposited, eat gine des mâles chez les abeilles sans aiguillon (Api- that egg, and oviposit themselves. In fact, dae, Meliponini), en particulier chez le genre Meli- sequences of up to 16 workers laying one after pona. On présente des arguments selon lesquels des facteurs écologiques, tels que la saison, combinés the other in the same cell have been observed. aux facteurs internes à la colonie, déterminent la Meanwhile, just a few cells away, a regular présence des mâles dans le temps. Les facteurs éco- POP may be occurring with queen oviposition logiques sont souvent si déterminants que les colo- and cell closure by a non-reproductive worker. nies d’une population sont hautement synchroni- The competing laying workers nearby are, this sées quant à leurs périodes de production de mâles et que la production de mâles est limitée à une sai- time, not interested in the cell containing the son donnée. Pourtant, sous certaines conditions tro- queen’s egg. This suggests that some cells are picales, ce synchronisme se perd et, au niveau de la protected from being used by workers, while population, des mâles sont produits tout au long de others are not. Perhaps the queen protects a cell l’année. Néanmoins chaque colonie ne les produit by adding a larger amount of a pheromone? By que dans certaines conditions et des lots de mâles varying this amount in a series of cells, the apparaissent dans des périodes distinctes de produc- tion de mâles (Fig. 1). Quand la production de queen may concentrate worker egg laying and mâles se fait aux dépens de la production d’ouvriè- egg destruction in some of the cells. This would res, les périodes de production de mâles peuvent protect a number of her own eggs and simulta- conduire à des fluctuations dans le nombre neously cause damage to worker reproduction. d’ouvrières et l’on suggère que ce facteur démogra- During the observed MPP, the mortality rates phique est l’un des facteurs de la régulation de la production de mâles. Un autre facteur est l’exis- for reproductive worker eggs were twice as tence d’une disparité entre le taux de construction high as those for queen eggs (Velthuis et al., des cellules d’ouvrières et le taux de production 2002). d’œufs de la reine. 182 H.H.W. Velthuis et al.

La production des mâles est indépendante des Männchen bei Stachellosen Bienen, mit besonderer mécanismes de production des reines. Les mâles Berücksichtigung der Melipona. Es werden Argu- constituent un moyen pour transférer des gènes à mente aufgezeigt, dass Umweltfaktoren wie Saiso- une autre colonie, alors que les reines servent soit à nalität in Kombination von Bedingungen im Volk multiplier les colonies par essaimage, soit à les das zeitliche Muster im Auftreten der Drohnen be- maintenir par remplacement de la reine. Chez six stimmen. Häufig dominieren die Umweltfaktoren, espèces bien étudiées de Melipona, il existe des dif- sodass die Drohnenerzeugung in den Völkern einer férences énormes en ce qui concerne l’investisse- Population stark synchronisiert wird. In diesen ment dans la reproduction. Ces différences portent Fällen gibt es Männchen nur zu festgelegten Zeiten. à la fois sur le sex-ratio primaire et sur l’investisse- Unter bestimmten tropischen Bedingungen geht ment total en sexués par rapport à l’investissement diese Synchronisation jedoch verloren und in Bezug en ouvrières (Tab. I). Pour chaque espèce le nombre auf die Population werden das ganze Jahr über Droh- de mâles fluctue beaucoup plus d’un mois à l’autre nen erzeugt. Trotzdem gibt es auf der Volksebene que le nombre de reines. nur unter speziellen Bedingungen Männchen; in so Les mâles proviennent des œufs pondus par les einem Volk treten Männchen nur während relativ ouvrières et aussi des œufs haploïdes pondus par la kurzen Perioden auf (Abb. 1). Weil in einem Teil reine. Chez certaines espèces la reine et l’ouvrière der Zellen Männchen statt Arbeiterinnen produziert reproductrice sont en compétition directe pour la werden, führt diese Periode zu einer Fluktuation in possession de la cellule dans laquelle elles peuvent der Anzahl der Arbeiterinnen. Es wird angenom- pondre leur œuf : un seul œuf est pondu par cellule. men, dass dieser demographische Faktor einer der Chez d’autres espèces pourtant, la compétition a Komponenten zur Regulation der Männchenpro- lieu entre les larves de la descendance, quand à la duktion ist. Ein weiterer Faktor ist das Auftreten fois la reine et l’ouvrière ont pondu un œuf dans la einer Diskrepanz bei der Rate beim Zellbau im Ver- cellule. De plus, les ouvrières peuvent détruire les gleich zur Rate der Eiproduktion der Königin. œufs de la reine, ou des autres ouvrières, avant de Die Drohnenerzeugung ist unabhängig vom Mecha- déposer elles-mêmes leur œuf. Et finalement une nismus der Königinnenproduktion. Männchen die- reine peut réagir à la ponte des ouvrières en pondant nen der Übertragung von Genen auf andere Völker, elle-même des œufs haploïdes, plutôt que diploïdes. während Königinnen entweder zur Vermehrung der Il existe donc de nombreuses façons par lesquelles Völker durch Schwärme beitragen, oder zum Erhalt les deux castes femelles entrent en compétition pour des Volkes als Nachfolgerinnen ihrer Mütter. Bei leur propre fitness. C’est le résultat d’interactions sechs gut untersuchten Arten von Melipona treten comportementales complexes pendant le processus deutliche Unterschiede in Bezug auf die Investition d’approvisionnement de la cellule et de ponte, qui in die Reproduktion auf. Diese Unterschiede betref- sont caractéristiques des abeilles sans aiguillon. fen beides, sowohl das primäre Geschlechtsverhält- Même au sein du genre Melipona, il existe des nis als auch die Gesamtinvestition in Geschlechts- différences importantes entre les espèces concer- tiere im Vergleich zur Investition in Arbeiterinnen nant le moment où cette compétition reine/ouvrière (Tab. I). Bei jeder Art ändert sich die Zahl der s’exprime (Fig. 2 ; Tab. II). Cela pourrait expliquer Männchen während des Jahres viel stärker als die les résultats contrastés des études portant sur la Zahl der Königinnen. paternité des mâles parmi les abeilles sans Drohnen können aus Eiern von Arbeiterinnen oder aiguillon. aus haploiden Eiern der Königin entstehen. Bei Les sex-ratios primaires des abeilles du genre Meli- einigen Arten, bei denen es pro Zelle nur ein Ei gibt, pona sont proches de l’unité. Il s’agit d’une valeur konkurrieren Königin und die reproduktiven remarquable pour un insecte social qui se multiplie Arbeiterinnen direkt um eine Zelle. Bei anderen par essaimage. On suppose qu’il s’agit de l’adapta- Arten jedoch findet die Konkurrenz zwischen den tion, au niveau de la colonie, au fait que les sites de Larven statt, nachdem sowohl die Königin als auch nidification libres sont très rares et que les colonies die Arbeiterin ein Ei in die Zelle gelegt haben. doivent se maintenir prêtes à essaimer sur un longue Außerdem kommt es vor, dass Arbeiterinnen das Ei période. der Königin oder anderer Arbeiterinnen vernichten, Les sex-ratios secondaires sont par contre très asy- bevor sie ihr eigenes Ei in die Zelle legen. métriques. On s’attend à ce qu’ils soient de l’ordre Schließlich ist es auch möglich, dass eine Königin de 1–0,1 % de reines/mâles. auf eierlegende Arbeiterinnen durch Ablage eigener haploider Eiern reagiert. Es gibt also viele verschie- abeille sans aiguillon / sex ratio / conflit reine- dene Möglichkeiten, wie die beiden weiblichen ouvrière / rassemblement de mâles / processus Kasten um ihre individuelle Fitness konkurrieren. d’approvisionnement et de ponte / cycle évolutif Das Ergebnis sind komplexe Verhaltensweisen und Interaktionen während der Verproviantierung der Zelle und des Ablaufs der Eiablage, ein sehr charak- Zusammenfassung – Drohnen der Melipona und teristischer Vorgang bei Stachellosen Bienen. anderer stachellosen Bienen und ihre Mütter. Selbst innerhalb der Gattung der Melipona gibt es Diese Arbeit beschäftigt sich vor allem mit dem wichtige Unterschiede zwischen den Arten in Produktionsmuster und der Abstammung der Bezug auf den Moment in dem der Königin- / Males of stingless bees and their mothers 183

Arbeiterin- Konflikt stattfindet. (Abb. 2, Tab. II). in Frieseomelitta varia worker bees, J. Morphol. Hiermit könnten sich die widersprechenden Ergeb- 242, 271–282. nisse bei Untersuchungen über die Abstammung Buren N.W.M., van, Sommeijer M.J. (1988) Étude der Männchen bei den Stachellosen Bienen des facteurs déterminants pour la domination erklären. reproductive de la Reine de Melipona trinitatis, Das ursprüngliche Geschlechtsverhältnis bei Meli- Act. Coll. Insectes Soc. 4, 285–290. pona liegt nahe bei 1:1, ein Wert, der bemer- Cameron E.C., Franck P., Oldroyd B.P. (2004) kenswert für soziale Bienen ist, die sich durch Genetic structure of nest aggregations and Schwärme vermehren. Es wird angenommen, dass congregations of the southeast Asian stingless dies eine Adaptation auf dem Volksniveau auf bee Trigona collina, Mol. Ecol. 13, 2357– Grund von selten vorhandenen Nistplätzen ist. Aus 2364. diesem Grund müssen die Völker über lange Camillo C. (1971) Estudos adicionais sobre os Zeiträume jederzeit in der Lage sein zu schwärmen. zangões de Trigona (Friesella) schrottkyi (Hym., Sie ziehen jederzeit junge Königinnen heran, die ), Ciênc. Cult. 23 (Suppl.), 273. meistens von ihnen getötet werden, sowie sie adult Camillo-Atique C. (1977) Estudo de variabilidade sind, sofern keine neue Nistmöglichkeit entdeckt etológica de Friesella schrottkyi incluindo a cara- wurde. Sekundäre Geschlechtsverhältnisse sind terização de espécies crípticas (Hymenoptera, jedoch sehr asymmetrisch. Sie liegen wahrschein- Meliponinae), These de Doutorado, Fac. Med. lich im Bereich von 1–0,1 % Königinnen im Ver- Ribeirão Preto. gleich zu Drohnen. Chinh T.X. (2004) Reproduction in social bees (Apidae: Apini: Meliponini), Ph.D. Thesis, Univ. Stachellose Bienen / Drohnen / Geschlechtsver- Utrecht. hältnis / life history / Drohnenansammlungen / Chinh T.X., Grob G.B.J., Meeuwsen F.J.A.J., Königinnen / Arbeiterinnen-Konflikt Sommeijer M.J. (2003) Patterns of male production in the stingless bee Melipona favosa (Apidae, Meliponini), Apidologie 34, 161– 170. REFERENCES Contel E.P.B., Kerr W.E. (1976) Origin of males in estimated from data of an Akahira Y., Sakagami S.F., Zucchi R. (1970) Die isozymic polymorphic system, Genética 46, 271– Nähreier von Arbeiterinnen einer stachellosen 277. Biene, Trigona (Scaptotrigona) postica, die von der Königin kurz vor der eigenen Eiablage Cortopassi-Laurino M. (1979) Observações sobre gefressen werden, Zool. Anz. 185, 85–93. atividades de machos de Plebeia droryana Friese (Apidae, Meliponinae), Rev. Bras. Entomol. 24, Alves D.A. (2004) Produção de operárias, rainhas e 177–191. machos em Melipona bicolor Lepeletier, 1836 (Apidae, Meliponini), Monografia Univ. Presbit. Cruz-Landim C. da, Cruz-Höfling M.A. (1971) Mackenzie, São Paulo. Cytochemical and ultrastructural studies on eggs from workers and queen of Trigona, Rev. Bras. Alves D.A., Ribeiro M.F., Santos Filho P.S., Pesq. Méd. Biol. 4, 19–26. Imperatriz-Fonseca V.L. (2004) Production of gynes and males in Plebeia remota Holmberg, Engels E., Engels W. (1984) Drohnenansammlungen 1903 (Apidae, Meliponini), in: Proc. 8th IBRA bei Nestern der stachellosen Biene Scaptotrigona Int. Conf. on Tropical Bees and VI Encontro postica, Apidologie 15, 315–328. sobre Abelhas, Ribeirão Preto, Funpec-editora, Engels W., Imperatriz-Fonseca V.L. (1990) Caste p. 753. development, reproductive strategies and control Bego L.R. (1982) On social regulation in of fertility in bees and stingless bees, in: Nannotrigona (Scaptotrigona) postica Latreille, Engels W. (Ed.), Social Insects, an evolutionary with special reference to male production cycles Approach to Caste and Reproduction, Springer (Hym., Apidae, Meliponinae), Bolm. Zool., Univ. Verlag, Berlin, pp. 167–230. São Paulo 7, 181–196. Faustino C.D., Silva-Matos E.V., Mateus S., Zucchi Bego L.R. (1990) On social regulation in R. (2002) First record of emergency queen Nannotrigona (Scaptotrigona) postica Latreille, rearing in stingless bees (Hymenoptera, , with special reference to productivity of colonies Meliponini), Insectes Soc. 49, 111–113. (Hymenoptera, Apidae, Meliponinae), Rev. Bras. Fisher R.A. (1930) The Genetic Theory of Natural Entomol. 34, 721–738. Selection, Clarendon Press, Oxford. Beig D. (1972) The production of males in queenright Grosso A.F., Bego L.R., Martinez A.S. (2000) The colonies of Trigona (Scaptotrigona) postica, J. production of males in queenright colonies of Apic. Res. 11, 33–39. Tetragonisca angustula (Hymenoptera, Melipon- Bezerra J.M.D. (1995) Aspectos da reprodução de inae), Sociobiology 35, 475–498. (Hymenoptera, Apidae), Imperatriz-Fonseca V.L. (1973) Miscellaneous These de Mestrado, Univ. Fed. Viçosa. observations on the behaviour of Schwarziana Boleli I.C., Simões Z.L.P., Bitondi M.M.G. (1999) quadripunctata, Bol. Zool. Biol. Marinha, Univ. Cell death in ovarioles causes permanent sterility São Paulo 30, 633–640. 184 H.H.W. Velthuis et al.

Imperatriz-Fonseca V.L., Zucchi R. (1995) Virgin Michener C.D. (1974) The Social Behavior of the queens in stingless bee (Apidae, Meliponinae) Bees, A comparative Study, Belknap Press, colonies: a review, Apidologie 26, 231–244. Cambridge, Mass. Imperatriz-Fonseca V.L., Kleinert A.M.P. (1998) Michener C.D. (2000) The Bees of the World, The Worker reproduction in a stingless bee Friesella Johns Hopkins University Press, Baltimore, schrottkyi (Apidae, Meliponinae), Entomol. Gen. Maryland. 23, 169–175. Michener C.D., Grimaldi D.A. (1988) A Trigona Imperatriz-Fonseca V.L., Matos E.T., Ferreira F., from late Cretaceous amber of New Jersey, Am. Velthuis H.H.W. (1998) A case of multiple Mus. Nov. 2917, 1–10. mating in stingless bees (Meliponinae), Insectes Moo-Valle H., Quezada-Euán J.J.G., Wenseleers T. Soc. 45, 231–233. (2001) The effect of food reserves on the Kerr W.E. (1950) Genetic determination of castes in production of sexual offspring in the stingless bee the genus Melipona, Genetics 35, 143–152. (Apidae, Meliponini), Insectes Kerr W.E. (1990) Why are workers in social Soc. 48, 398–403. Hymenoptera not males? Rev. Bras. Genét. 13, Nogueira-Ferreira F.H., Soares A.E.E. (1998) Male 133–136. aggregations and mating flight in Tetragonisca Koedam D. (1999) Production of queens, workers and angustula (Hymenoptera, Apidae, Meliponinae), males in the stingless bee Melipona favosa Iheringia Sér. Zool., Porto Alegre, RS 84, 141– (Apidae: Meliponinae): patterns in time and 144. space, Neth. J. Zool. 49, 289–302. Nogueira-Neto P. (1997) Vida e Criação de Abelhas Koedam D., Aguilar Monge I., Sommeijer M.J. Indígenas sem Ferrão, Nogueirapis, São Paulo. (1995) Social interactions of gynes and their Nonacs P., Carlin N.F. (1990) When can ants longevity in queenright colonies of Melipona discriminate the sex of brood? A new aspect of favosa (Apidae, Meliponinae), Neth. J. Zool. 45, queen-worker conflict, Proc. Natl. Acad. Sci. 480–494. (USA) 87, 9670–9673. Koedam D., Velthausz P.H., Krift T. van der, Dohmen Paxton R.J. (2000) Genetic structures of colonies and M.R. (1996) Morphology of reproductive and male aggregations in the stingless bee Scaptotri- trophic eggs and their controlled release by work- gona postica, as revealed by microsatellite analy- ers in Trigona (Tetragonisca) angustula Illiger sis, Insectes Soc. 47, 63–69. (Apidae, Meliponinae), Physiol. Entomol. 21, Paxton R.J. (2005) Male mating behaviour and 289–296. mating systems of bees: an overview, Apidologie Koedam D., Contrera F.A.L., Imperatriz-Fonseca 36, 145–156. V.L. (1999) Clustered male production by workers in the stingless bee Melipona subnitida Paxton R.J., Ruhnke H., Shah M., Bego L.R., Ducke (Apidae, Meliponinae), Insectes Soc. 46, Quezada-Euán J.J.G., Ratnieks F.L.W. (2001) 387–391. Social evolution in stingless bees: are the workers or is the queen in control of male parentage? in: Koedam D., Velthuis H.H.W., Dohmen M.R., Quezada-Euan J.J.G. (Ed.), II Seminario Imperatriz-Fonseca V.L. (2001) The behaviour of mexicano sobre abejas sin aguijón, Univ. laying workers and the morphology and viability Autónoma de Yucatan, Mérida, México, pp. 104– of their eggs in Melipona bicolor bicolor, 107. Physiol. Entomol. 26, 254–259. Paxton R.J., Bego L.R., Shah M.M., Mateus S. (2003) Koedam D., Imperatriz-Fonseca V.L. (2004a) The Low mating frequency of queens in the stingless sharing of male production among worker cohorts bee Scaptotrigona postica and worker maternity in Melipona (Apidae, Meliponini), in: Proc. 8th of males, Behav. Ecol. Sociobiol. 53, 174–181. IBRA Int. Conf. Tropical Bees and VI Encontro sobre Abelhas, Ribeirão Preto, Funpec-editora, Peters J.M., Queller D.C., Imperatriz-Fonseca V.L., pp. 264–268. Roubik D.W., Strassmann J.E., (1999) Mate number, kin selection, and social conflicts in Koedam D., Imperatriz-Fonseca V.L. (2004b) Cell stingless bees and honey bees, Proc. R. Soc. sealing and egg protection in Melipona bicolor Lond. B 266, 379–384. (Apidae, Meliponini), in: Proc. 8th IBRA Int. Conf. on Tropical Bees and VI Encontro sobre Ratnieks F.L.W. (2001) Heirs and spares: caste Abelhas, Ribeirão Preto, Funpec-editora, p. 767. conflict and excess queen production in Melipona bees, Behav. Ecol. Sociobiol. 50, 467–473. Koedam D., Contrera F.A.L., Fidalgo A. de O., Imperatiz-Fonseca. V.L. (2005) How queens and Ribeiro M.F., Alves D.A. (2001) Size variation in workers share in male production in the stingless Schwarziana quadripunctata (Holmberg) queens bee Melipona subnitida Ducke (Apidae, (Hymenoptera, Apidae), Rev. Etol. São Paulo 3, Meliponini), Insectes Soc. (in press). 59–65. Lacerda L.M., Zucchi R. (1999) Behavioral Roubik D.W. (1989) Ecology and Natural History of alterations and related aspects in queenless Tropical Bees, Cambridge Univ. Press, Cam- colonies of Geotrigona mombuca (Hymenoptera, bridge, Mass. Apidae, Meliponinae), Sociobiology 33, 277– Roubik D.W. (1990) Mate location and mate compe- 288. tition in males of stingless bees (Hymenoptera: Males of stingless bees and their mothers 185

Apidae: Meliponinae), Entomol. Gen. 15, 115– A., Arnold G., Lepage M. (Eds.), Les Insectes 120. Sociaux, Université Paris Nord, Paris. Sakagami S.F. (1982) Stingless bees, in: Hermann Terada Y. (1974) Contribuição ao estudo da regulação H.R. (Ed.), Social Insects, Academic Press, New social em muelleri e Frieseomelitta York, Vol. 3, pp. 361–423. varia (Hymenoptera, Apidae), These de Mestrado, Sakagami S.F., Zucchi R. (1968) Oviposition Universidade de São Paulo, São Paulo, 96 p. behavior of an Amazonic stingless bee, Trigona Thornhill R., Alcock J., (1983) The Evolution of (Duckeola) ghilianii, J. Fac. Sci. Hokkaido Univ. Insect Mating Systems, Harvard University Press, Zool. 16, 564–581. Cambridge, Massachusetts. Sakagami S.F., Zucchi R. (1974) Oviposition Tóth E., Queller D.C., Imperatriz-Fonseca V.L., behavior of two dwarf stingless bees Hypotrigona Strassmann J.E. (2002) Genetic and behavioral (Leurotrigona) muelleri and H. (Trigonisca) conflict over male production between workers duckei with notes on the temporal articulation of and queens in the stingless bee Paratrigona oviposition process in stingless bees, J. Fac. Sci. subnuda, Behav. Ecol. Sociobiol. 53, 1–8. Hokkaido Univ. Zool. 19, 361–421. Tóth E., Queller D.C., Dollin A., Strassmann J.E. Silva D.L.N. da (1977) Estudos bionômicas em (2004) Conflict over male parentage in stingless colônias mistas de Meliponinae (Hymenoptera, bees, Insectes Soc. 51, 1–11. Apoidea), Bol. Zool. Univ. São Paulo 2, 7– Trivers R.L., Hare H. (1976) Haplodiploidy and the 106. evolution of the social insects, Science 191, 249– Silva D.L.N. da, Zucchi R., Kerr W.E. (1972) 276. Biological and behavioral aspects of the Veen J.W. van, Sommeijer M.J., Meeuwsen F. (1997) reproduction in some species of Melipona Behaviour of drones in Melipona (Apidae, (Hymenoptera, Apidae, Meliponinae), Anim. Meliponinae), Insectes Soc. 44, 435–447. Behav. 20, 123–132. Veen J.W. van, Sommeijer M.J. (2000) Observations Silva-Matos E.V., Noll F.B., Zucchi R. (2000) on gynes and drones around nuptial flights in the Sistemas de regulação social encontrados em stingless bees Tetragonisca angustula and abelhas altamente eussociais (Hymenoptera; Melipona beecheii (Hymenoptera, Apidae, Apidae, Meliponinae), Anais IV Encontro sobre Meliponinae), Apidologie 31, 47–54. Abelhas, Ribeirão Preto, pp. 95–101. Velthuis H.H.W. (1993) Pollen digestion and the Singer R.B., Flach A., Koehler S., Marsaioli A.J., evolution of sociality in bees, Bee World 73, 77– Amaral M.C.E. (2004) Sexual in Mormo- 89. lyca ringens (Lindl.) Schltr. (: Max- illariinae), Ann. Bot. 93, 755–762. Velthuis H.H.W., Sommeijer M.J. (1991) Roles of Sommeijer M.J., Buren N.J.M. van (1992) Male morphogenetic hormones in caste polymorphism production by laying workers in queenright in stingless bees, in: Gupta A.P. (Ed.), colonies of Melipona favosa (Apidae, Morphogenetic Hormones of , Meliponinae), in: Billen J. (Ed.), Biology and Rutgers Univ. Press, New Jersey, pp. 346–383. Evolution of Social Insects, Leuven Univ. Press, Velthuis H.H.W., Araujo Alves D. de, Imperatriz- Leuven, pp. 89–97. Fonseca V.L., Duchateau M.J. (2002) Worker Sommeijer M.J., Bruijn L.L.M. de (1995) Drone bees and the fate of their eggs, Proc. Exp. Appl. congregations apart from the nest in Melipona Entomol., NEV Amsterdam 13, 97–102. favosa, Insectes Soc. 42, 123–127. Wenseleers T., Ratnieks F.L.W. (2004) Tragedy of Sommeijer M.J., Chinh T.X., Meeuwsen F.J.A.J. the commons in Melipona bees, Proc. R. Soc. (1999) Behavioural data on the production of Lond. B, Suppl. 5, S310–312. males by workers in the stingless bee Melipona Wenseleers T., Hart A.G., Ratnieks F.L.W., Quezada- favosa (Apidae, Meliponinae), Insectes Soc. 46, Euán J.J.G. (2004) Queen execution and caste 92–93. conflict in the stingless bee Melipona beecheii, Sommeijer M.J., Bruijn L.L.M. de, Meeuwsen Ethology 110, 725–736. F.J.A.M., Martens E.P. (2003) Natural patterns of Zucchi R. (1993) Ritualized dominance, evolution of caste and sex allocation in the stingless bees queen worker interactions and related aspects in Melipona favosa and M. trinitatis related to stingless bees (Hymenoptera: Apidae), in: Inoue worker behaviour, Insectes Soc. 50, 38–44. T., Yamane S. (Eds.), Evolution of Insect Suka T., Inoue T. (1993) Nestmate recognition of societies. Hakushinsha, Tokyo, pp. 207–249. the stingless bee Trigona (Tetragonula) minang- Zucchi R., Silva-Matos E.V., Nogueira-Ferreira F.H., kabau (Apidae: Meliponinae), J. Ethol. 11, 141– Azevedo G.G. (1999) On the cell provisioning and 147. oviposition process (POP) of the stingless bees, Suka T., Inoue T., Roubik D.W. (1994) Worker nomenclature reappraisal and evolutionary con- ovipostion and kin recognition of the stingless siderations (Hymenoptera, Apidae, Meliponinae), bee Scaptotrigona barrocoloradensis, in: Lenoir Sociobiology 34, 65–86.