Content uploaded by Marinus J. Sommeijer
Author content
All content in this area was uploaded by Marinus J. Sommeijer
Content may be subject to copyright.
Summary. This paper is the report of a long-duration study
of the production of gynes and males in Melipona stingless
bees in Trinidad and Tobago, including a behavioural study of
laying workers. In 167 brood samples (16342 pupae) of M.
favosa, taken from 78 colonies in the natural environment
over the years 1993–2001, workers represented 78.4% and
males represented 17.3%. Gynes represented 5.1% of all fe-
male brood. Gyne proportion never exceeded 22.5%. Gynes
were found to be common throughout the year. Like the gy-
nes, males were present all year, but their occurrence varied
over the months. Males were most common in the months of
July and August. At the colony level male occurrence in the
brood was characterized by very high values (maximum
74.2%), next to a frequent complete absence (in 26.5% of all
samples). The finding of bout-like production of males at the
colony level is supported by the observations on reproductive
worker behaviour. The colonies appeared to be in different
phases of male production. When male producing laying
workers were present, they competed heavily for egg-laying
opportunities. Various behaviours of laying workers and the
queen are described and interpreted on the basis of repro-
ductive competition and kin conflict theories. From these re-
sults, we conclude that in M. favosa, workers have a distinct
influence on reproduction, including the production of gynes
and males.
Key words: Male production, gyne production, worker be-
haviour, stingless bees.
Introduction
In eusocial Hymenoptera the mated female has the possibil-
ity to decide whether she will lay fertilised or unfertilised
eggs, resulting in female and male brood respectively. In some
taxa unmated workers may participate in the production of
males (Crozier and Pamilo, 1996; Bourke and Franks, 1995;
Bourke and Ratnieks, 1999). In species where colonies are
founded by single females, usually many sexuals are pro-
duced. If colony multiplication takes place by ‘swarming’,
sexual females are usually produced in small numbers and
only when needed. Males are seasonally numerous when
swarming is seasonal. When swarming can take place over the
entire year, one would expect males and sexual females at the
population level to be present all year. However, Melipona
stingless bees have a very low incidence of ‘swarming’.
Colonies of these species have a low mortality and laying
queens have a high longevity. This raises questions concern-
ing the production of sexuals in this group.
Diploid brood in eusocial Apidae can develop into two
different castes: workers or queens. In most eusocial bees,
queens emerge from ‘queen cells’ that are distinctly larger
than cells from which workers emerge. In these species, queen
determination is trophically regulated through larval food
(Michener, 1976). Melipona bees are unique for the rearing
of all castes and sexes from a single-sized brood cell. In
Melipona, queens are produced continuously and in high
numbers (Sakagami, 1982). In this genus queen determina-
tion has been proposed to have a genetic component in addi-
tion to being influenced by trophic factors (Kerr et al., 1962;
Velthuis and Sommeijer, 1990; 1991). Recently, the high pro-
duction of queens in Melipona has been explained through the
occurrence of a caste conflict (Ratnieks, 2001). The haplo-
diploid system determines different degrees of relatedness of
female bees with their own offspring as compared to that with
their sisters. The proposed caste conflict hypothesis implies
that female Melipona larvae, through self-determination, tend
to develop into queens whereas workers tend to produce sis-
ter workers.
This article contains the results of our long-term study on
caste and sex allocation in M. favosa. Focussing on the spe-
cial role of workers in the production of the different types of
Insectes Soc. 50 (2003) 38–44
0020-1812/03/010038-07
© Birkhäuser Verlag, Basel, 2003
Insectes Sociaux
Research article
Natural patterns of caste and sex allocation in the stingless bees
Melipona favosa and M. trinitatis related to worker behaviour
M.J. Sommeijer1,L.L.M.de Bruijn1,F.J.A.J.Meeuwsen1and E.P. Martens2
1Utrecht University, Social Insects Department, P.O. Box 80.086, NL-3508 TB Utrecht, The Netherlands, e-mail: m.j.sommeijer@bio.uu.nl
2Utrecht University, Centre of Biostatistics, Padualaan 14, NL-3584 CH Utrecht, The Netherlands, e-mail: e.p.martens@bio.uu.nl
Received 12 September 2001; revised 15 July 2002; accepted 21 August 2002.
Insectes Soc. Vol. 50, 2003 Research article 39
months. However, the relative occurrence of gynes at the pop-
ulation level did not vary significantly in relation to the time
of the year (Kruskal-Wallis, p = 0.599).
The occurrence of males in M. favosa
Males were also found during all months of the year. The
monthly occurrence of males at the population level is given
in Figure 3. In contrast with the occurence of gynes their
numbers differed significantly over months (Kruskal-Wallis,
P<0.001), reaching peak values in the months of July and
August (25.9% and 29.7% respectively). Minimum values
were found in samples of May and in the (small number) of
individuals in the colony, we will also describe in detail the
behaviour and particularly the social interactions of drone-
producing workers. The occurrence of such worker behaviour
in a series of colonies under simultaneous observation will
be analysed. The theoretical impact of the results will
be discussed at the colony level as well as at the population
level.
Materials and methods
In the period of 1993–2001, we analysed 176 brood samples (contain-
ing a total number of 16342 brood cells) from 78 different colonies of
M. favosa in Trinidad and Tobago, W.I. These colonies were located at
different sites in their natural habitat. Multiple samples of the same
colonies were not taken within 6 months of each other to ensure inde-
pendance of the samples. The samples were taken throughout the year
from brood containing pupae of an age at which they have at least
coloured eyes. Whenever possible, we extracted entire combs as a sam-
ple. Some samples were immediately analysed for sex and caste, others
were preserved in alcohol and analysed at a later moment. For the analy-
ses we discarded the samples which consisted of less than 40 cells.
Of M. trinitatis, 21 samples were analysed, taken in four different
months from 15 colonies at two different locations.
Oviposition behaviour was studied during three observations periods
of 29, 14 and 15 hours respectively, all within one month in 1998. In this
way, we simultaneously recorded the behaviour of workers during ovipo-
sition in ten colonies of M. favosa, some of which were producing males.
The colonies had populations of 100 to 400 bees, which is normal for
thriving colonies of this species. They were housed in hives with a glass
lid and installed in their natural habitat where they could forage nor-
mally. Through the non-stop intensive recordings by four experienced
observers it was possible to observe all ‘Provisioning and Oviposition
Processes, POP’s’ (Sakagami, 1982) during those days. Special attention
was paid to the behaviour and social interactions of the workers and the
queen during each POP.
Results
The occurrence of castes and sexes
Of the total number of pupae of M. favosa workers repre-
sented 78.4% and males represented 17.3%. Gynes repre-
sented 5.1% of all female brood. For the smaller data set on
M. trinitatis (2567 pupae), these figures were 87.6%, 7.7%
and 4.8% respectively.
The occurrence of gynes in M. favosa
At the level of the samples, there was variation in the per-
centage of gynes, and in most of the samples gynes were pre-
sent. In Figure 1a frequency distribution is plotted for the
gyne-proportion of females. Although we found 14% of the
samples to contain no gynes at all, at the population level gy-
nes were found to occur throughout the year. The proportion
gynes-of-females ranged from 0 to 22.5% (average: 5.0%; sd:
4.4%). The monthly occurrence of gynes, as a percentage of
all females is given in Figure 2. We have no samples for the
months of April and June, whereas the numbers of samples
for March and October are small compared to those in other
Figure 1. Frequency distribution of the proportion of gynes-of-females
in brood samples of Melipona favosa (n = 176 samples) and Melipona
trinitatis (n = 21 samples)
40 M.J. Sommeijer et al. Caste and sex allocation and worker behaviour in Melipona
Tab le 1. POP-rates per colony during three observation series
Colony: T1 T2 T7 B1 B2 W2 B3 B5 W6 W8 all colls.
# POP’s 25–27 Jul (29 hrs) 22 31 22 4 19 6 104
pop/hr 0.76 1.07 0.76 0.14 0.66 0.21 0.6
# POP’s 6 Aug (14 hrs) 6 10 12 12 7 9 9 19 13 97
pop/hr 0.43 0.71 0.86 0.86 0.5 0.64 0.64 1.36 0.93 0.77
# POP’s 14 Aug (15 hrs) 19 1 17 16 17 13 14 18 11 126
pop/hr 1.27 0.07 1.13 1.07 1.13 0.87 0.93 1.2 0.73 0.93
total # POP’s 47 42 51 32 43 6 22 23 37 24 327
average POP-rate 0.81 0.72 0.88 0.55 0.74 0.21 0.76 0.79 1.28 0.83 0.75
Figure 2. Monthly occurrence of gynes at the population level Figure 3. Monthly occurrence of males at the population level
1We tested whether mutiple samples over the period of eight years taken
from some of the colonies were indeed independant by randomly
selecting only one sample per colony. Differences in male occurrence
between months remained significant.
Occurrence of ovipositions in M. favosa
Oviposition rates
During the three intranidal observation periods in 10 different
colonies a total number of 327 POP’s was observed. The
results are presented in Table 1. The oviposition rate for all
colonies during the total observation time of 58 hours aver-
aged at 0.8 POP’s/colony/hour. Individual colonies had dif-
ferent POP-rates for the total observation time, ranging from
0.2 – 1.3 POP’s/hour. The table indicates how POP rate can
also fluctuate within colonies.
The frequency of different POP-types
Oviposition behaviour occurred in different forms. We dis-
criminate between two major forms: Queen-POP’s and
Worker-POP’s (17.1% of all POP’s). In Queen-POP’s, the egg
that is finally operculated is laid by the queen, and in Worker-
POP’s this egg is laid by a worker. Worker-POP’s are charac-
samples of October (3.1% and 2.3% respectively).1At the
colony level, males occured in 73.5% of the samples and their
occurrence in the brood is characterized by very high values
(maximum 74.2%), next to a frequent complete absence.
The occurrence of sexuals in M. trinitatis
In M. trinitatis gynes and males were found to occur in sim-
ilar patterns as in M. favosa. Gynes were found to occur in
80.0% of all samples (n= 20), and gyne percentages in our
samples never exceeded 15.9% of all females (Fig. 1b).
Males however were often absent. They only occurred in
45.0% of the samples. When males were present, their per-
centage could reach high values (up to 30.7% of the pupae in
the sample).
Insectes Soc. Vol. 50, 2003 Research article 41
Behaviour of workers and the queen during POP’s in
colonies with RLW’s
The behaviour of queen and workers during a POP is very dif-
ferent between colonies where Worker-POP’s occur and
colonies that only have Queen-POP’s. The start of Worker-
POP’s is lengthy when the queen is not present at the very be-
ginning.
Description of worker behaviour during Worker-POP’s
At the start of a Worker-POP generally more workers are pre-
sent at that cell than in POP’s in colonies that are not in RLW-
phase (about 6 versus 2 workers). Workers in RLW-phase do
not perform the typical retreats from the queen (c.f. Som-
meijer and de Bruijn, 1984). In Worker-POP’s the food re-
gurgitations are usually temporally interrupted after the first
two or three discharges have taken place. During these inter-
ruptions workers may take up some of the larval food. The
competition between workers in these POP’s is clearly visible.
Workers try to push each other away from the cell and they
antennate each other very intensively. Worker interference in-
creases toward the end of a Worker-POP.
In this study we confirmed the earlier described behav-
ioural features of RLW’s. Particularly obvious in these obser-
vations was the strong drive of RLW’s to remain sitting on the
cell in which they had oviposited, in order to proceed imme-
diately with the operculation of the cell. This is in very sharp
contrast to the behaviour of laying workers that release trophic
worker eggs. Such workers immediately depart from the cell
after egg laying. The occurrence of various laying workers in
a single Worker-POP is always characterised by aggressive
mutual interactions. For example, a worker that assumes the
laying position may be interfered by another worker that fi-
nally obtains a position underneath the first worker, from
which position she can take over egg-laying. Workers may
also eat worker eggs and replace it by their own egg (in 7%
of the Worker-POP’s).
Description of queen behaviour during Worker-POP’s
The locomotive pattern of the queen in colonies with Worker-
POP’s is different from that in other colonies. In about ten
percent of the cases the queen was not even present during
the start of a Worker-POP. In fifty percent of the Worker-
POP’s the queen departed from the cell after the first food
discharges. These departures did not appear to be the result
of pushing or other interference by workers. In colonies
with Worker-POP’s we also observed the queen frequently
move to very distant parts of the nest, where she is not en-
countered in other colonies. During Worker-POP’s the queen
does not as heavily antennate and drum on discharging work-
ers with her front legs as is the case in colonies outside the
RLW-phase.
terized by the immediate operculation of the cell by the lay-
ing worker (a so-called ‘reproductive laying worker’RLW). In
each of these major types variation consists of whether or not
a trophic worker egg is released preceeding the laying of the
egg that is finally operculated. The frequencies of all four
POP-forms are listed in Table 2. Of all queen ovipositions,
28.8% was preceeded by the release of one or more trophic
worker eggs. For all reproductive worker eggs this was the
case in only 7.1%. In 327 observed POP’s we never recorded
a POP in which a worker egg was released after a queen’s
oviposition.
Comparing the oviposition rate of colonies in RLW-phase
with this in colonies without Worker-POP’s (Fig. 4), it can be
seen that the POP-rates in general are higher during RLW-
phase than out of RLW-phase. With the small sample sizes it
can not be shown that this difference is significant (Mann-
Whitney test: p= 0,079).
Tab le 2. The frequencies of the two major POP-forms (with a Queen
Egg, QE, or with a Reproductive Worker Egg, RWE) and of their sub-
forms (preceeded or not by a Trophic Worker Egg, TWE)
Frequency (%) % of total POP’s
TWE(s) + QE 78 (28.8) 23.9
QE 193 (71.2) 59
Total Queen-POP 271 82.9
TWE(s) + RWE 4 (7.1) 1.2
RWE 52 (92.9) 15.9
Total Worker-POP 56 17.1
Total POP’s 327 (100%)
Figure 4. Boxplots of oviposition rates during and out of Reproductive
Laying Worker (RLW)-phase
42 M.J. Sommeijer et al. Caste and sex allocation and worker behaviour in Melipona
Description of behaviour in Queen-POP’s in colonies in
RLW-phase
Queen-POP’s in colonies in RLW-phase are different from
POP’s that take place in colonies where RLW’s are absent (c.f.
Sommeijer and de Bruijn, 1984). They are more time-con-
suming through the little active behaviour of the queen and
through her frequent departures from the cell. In twenty per-
cent of these Queen-POP’s we observed the uptake of larval
food by the workers during provisioning. The eating of worker
eggs by sister workers prior to the final oviposition by the
queen occurred in two cases. These slowed-down POP’s could
lead to oviposition of four trophic eggs per POP. On the other
hand the reluctance of the queen to eat the trophic eggs even
resulted once in the presence of two worker eggs in the same
cell.
Discussion
Occurrence of gynes
A continuous queen production in Melipona colonies has ear-
lier been established (Kerr, 1950; Kerr et al., 1962; Ratnieks,
2001; Van Veen et al., 2000). Our present results again con-
firm that gynes are always present in colonies. In these sam-
ples we never obtained gyne-of-females percentages sur-
passing 22.5% for M. favosa and 15.9% for M. trinitatis. The
hypothesis that queen determination in Melipona is regulated
through the combination of genetic and nutritional determi-
nation factors (Kerr et al., 1962; Velthuis and Sommeijer,
1991; Sommeijer and De Bruijn, 1994) predicts maximum
gyne proportions of 25%. According to the recently proposed
caste conflict model by Ratnieks and Wenseleers (Ratnieks,
2001; Wenseleers et al., 2001), the optimum proportion of fe-
males that should develop into queens ranges from 14 to 20
percent. Our results provide support to both hypotheses. Con-
siderable variation in gyne proportion occurs in our samples
and such a large series of samples appears necessary to ac-
curately estimate the proportion of gynes. Spatial clustering
of gyne pupae in the combs, as reported by Koedam (1999),
should have lead to an occasional occurrence of a gyne per-
centage over 25% in at least some of our small samples.
Variation of gyne proportion in brood samples is not re-
flected at the population level. At this level gynes are present
in the same proportion throughout the year.
Male presence at the colony and the population level
Male proportion in the samples varies strongly. High per-
centages, next to complete absence of emerging males have
earlier been found in M. favosa by Chinh and co-workers
(Chinh et al., in press). These authors also established that the
typical pattern of rhythmic production of males in individual
colonies of M. favosa is the result of rhythmically occurring
reproductive laying worker behaviour in these colonies. It was
found that through this process a colony does not produce
males for periods of more than several months despite ongo-
ing brood cell production.
Within the population males are present throughout the
year, but distinct differences in male percentages are found
between months. A seasonal production of sexuals as a direct
result of changing environmental factors is not to be expected
in view of the fact that the development of young bees is a
slow process in this group of bees. This does not allow for a
fast reaction to these changes. However, the fluctuation in the
number of males over the year can possibly be explained by
the seasonally varying food stores of the colonies, as this has
been indicated to correlate with male production (Moo-Valle
et al., 2001).
Production of males by reproductive laying workers
In this study we confirmed the earlier finding (Som-
meijer et al., 1999) that reproductively laying workers are
common in M. favosa. It was also confirmed that the colonies
in a population may occur in different phases of RLW activ-
ity: some colonies may be male producing while other
colonies are not in this phase. Chinh and co-workers (in press)
demonstrated how in M. favosa RLW behaviour (in the
Reproductive Laying Worker Period, RLWP) is followed by
the Male Emergence Period, MEP. An ultimate function of
colonies occurring in different male producing phases may be
that this asynchrony prevents inbreeding. This mechanism
further garantees a continuous presence of males in the
population.
The different forms of worker oviposition
The behavioural elements during the different types of POP’s
appear to be similar to the descriptions of POP’s we earlier
published (Sommeijer and Van Buren, 1992). In the present
study we did not find the release of reproductive worker eggs
after the queen’s oviposition. Such a method for male pro-
duction, described to be normal in Scaptotrigona postica
(Bego, 1990; Beig, 1972) is certainly not a regular way for
male production in M. favosa. The release of reproductive
worker eggs after queen oviposition was only rarely seen by
us in a colony where the queen was to be replaced (Chinh
et al., in press).
Queen-worker relations and proximate mechanisms for the
occurrence of male producing workers
In many aspects Worker-POP’s in these queenright colonies
resemble POP’s in experimentally dequeened colonies, in
which there is also much aggression between workers (Som-
meijer and Velthuis, 1977). Overt mutual worker aggression
also occurs in colonies where the queen is recently accepted
and just starts to oviposit. In these colonies POP’s are also
characterised by a lengthy duration and by the queen leaving
the cell frequently during POP (De Bruijn pers. obs.). It ap-
Insectes Soc. Vol. 50, 2003 Research article 43
the presently studied Worker-POP’s, should also be inter-
preted to have a function in the reproductive competition
among workers.
Worker control over reproduction
In addition to their role in the production of males through the
laying of reproductive worker eggs, workers also may influ-
ence queen production. Their behaviour during POP serves to
regulate the quantity and quality of larval food in cells. Work-
ers also play a key role in worker production by regulating cell
construction. Typically in stingless bees, construction of a
brood cell is linked to the subsequent POP in that cell (Saka-
gami, 1982). The most important constructors of a brood cell
are, immediately after this, the provisioners of the same cell
(Sommeijer et al. 1982). Further, some of the building work-
ers initiate POP’s by urging the queen to visit the cell that is
ready to be oviposited (Sommeijer and de Bruijn, 1985). Dif-
ferent from the situation in Apis, worker control is evident for
many aspects of the intranidal and extranidal reproductive bi-
ology of Melipona stingless bees. As we are now able to point
to the evolutionary function of worker reproductive behav-
iours, study of their proximate mechanisms is urgently
needed.
Acknowledgement
Charles D. Michener and Mary-Jane West Eberhard gave very stimulat-
ing comments at the earlier version of this manuscript. Koos Boomsma
also read the previous version. Tom Wenseleers and an anonymous
reviewer supplied detailed suggestions for improvement. Frouke
Hofstede and Tong X. Chinh collaborated in collecting some of the data.
We are thankful to a number of persons in Trinidad and Tobago W.I. who
provided essential assistance for our studies. Gladstone Solomon, pres-
ident of the Tobago Apicultural Society and many other beekeepers on
the island of Tobago have much contributed to this work and have made
the field studies possible. We wish to thank the Tobago House of As-
sembly (Mr. Hugh McKennah, Secretary of Agriculture) and the com-
plete staff of the Agricultural Department at the Botanic Station, of
whom we particularly wish to mention Mr. Jerry Keens Dumas, Ad-
ministrator, and Mr. E. Harris, Chief Technical Officer. Mr. Edson
George, Bon Accord, Tobago, and Mr. Harrypersad Ramsamooj, Rio
Claro, Trinidad, supplied substantial support. Ms. Sheila McNab pro-
vided linguistic assistance.
References
Arce Arce H.G. and J.W. van Veen, 1994. Seasonal brood dynamics and
foodstorage in Melipona beecheii (Apidae: Meliponinae). In: Les In-
sectes Sociaux (A. Lenoir, G. Arnold and M. Lepage, Eds.), Publ.
Univ. Paris Nord p. 372.
Bego, L.R., 1990. On social regulation in Nannotrigona (Scaptotrigona)
postica Latreille, with special reference to productivity of colonies
(Hymenoptera, Apidae, Meliponinae). Revta. bras. Ent. 34:
721–738.
Beig, D., 1972. The production of males in queenright colonies of Trig-
ona (Scaptotrigona) postica. J. Apicult. Res. 11: 33 –39.
Bourke, A.F.G. and F.L.W. Ratnieks, 1999. Kin conflict over caste de-
termination in social Hymenoptera. Behav. Ecol. Sociobiol. 46:
287–297.
pears as if the queen loses her typical queen-like status in
a colony that enters RLW-phase. This could point to an in-
trinsic rhythmic process in the queen. However, because we
only saw RLW-phases in well populated, thriving colonies, we
hypothesise that the rhythmicity of the male producing
process by workers results from repeatedly occurring changes
in workers. During, or preceding, the RLW-phases some
workers develop into reproductives. These workers interact
differently with the queen and cause the behaviour of the
queen during Worker-POP’s as well as during Queen-POP’s in
RLW- phases to change. We assume that the development of
these workers is not directly triggered by food factors but in-
stead that these workers are programmed to become repro-
ductives and that this can only be expressed under favourable
colony conditions. The necessary colony conditions are as yet
not clear, but we consider colony population size an important
factor. In large colonies, where oviposition rate is high, it is
physically impossible for the queen to keep control over every
POP as much as in smaller colonies that have lower oviposi-
tion rates.
Ultimate explanation for the behaviour of reproductive
laying workers by kin conflict
The behavioural characteristics of workers that lay repro-
ductive eggs and compete with the queen or with other work-
ers (e.g. through mutual worker egg eating) are evidence for
the reproductive competition between these females. It is gen-
erally accepted that genetic relations within the colony of
haplo-diploid social Hymenoptera should lead to reproductive
competition (Bourke, 1994; Bourke and Franks, 1995;
Bourke and Ratnieks, 1999; Crozier and Pamilo, 1996). The
relatedness of a laying worker to sons of the queen is lower
than her relatedness to her own sons. In addition, sons of a
particular worker are genetically different from those pro-
duced by sister-workers (Ratnieks, 1988). The observed in-
tense behavioural competition between the various RLW’s
that oviposit in the same brood cell can be ultimately ex-
plained by these differences in relatedness. Even the hesita-
tion of workers to proceed with food discharges after the first
discharges have been made, should be interpreted as resulting
from kin conflict between reproductive workers. In the typi-
cal POP sequence of Melipona, only when a certain amount
of larval food has already been deposited in the cell, laying
workers will be able to perform the final discharge. Ex-
clusively after this final discharge, they can take up the lay-
ing posture on that cell. The observed behaviour of the
provisioners reflects a competition for being the final dis-
charger which allows access to the ovipositing position on
that cell. The earlier reported peculiar interruptions in the lar-
val food discharging sequence in queenless colonies should
be similarly explained. We have established that in queenless
colonies the actual laying worker very often is the first to
release some of the larval food. By doing so, she initiates
the sequence of discharges, which she can conclude by
her own oviposition (Sommeijer and Velthuis, 1977). Uptake
of larval food from the cell by a worker, again observed in
44 M.J. Sommeijer et al. Caste and sex allocation and worker behaviour in Melipona
Bourke, A.F.G. and N.R. Franks, 1995. Social Evolution in Ants. Prince-
ton University Press, New Jersey. 529 pp.
Bourke, A.F.G., 1994. Worker matricide in social bees and wasps. J.
Theor. Biol. 167: 283 – 292.
Chinh, T.X., G.B.J. Grob, F.J.A.J. Meeuwsen and M.J. Sommeyer, in
press. Patterns of male production in the stingless bee Melipona
favosa (Apidoe, Meliponini). Apidologie.
Contel, E.P.B. and W.E. Kerr, 1976. Origin of males in Melipona subni-
tida estimated from data of an isozymic polymorphic system. Ge-
netica 46: 271–279.
Crozier, R.H. and P. Pamilo, 1996. Evolution of Social Insect Colonies:
Sex Allocation and Kin Selection. Oxford University Press, Oxford.
306 pp.
Engels, W. and V.L. Imperatriz-Fonseca, 1990. Caste development, re-
productive strategies, and control of fertility in honey bees and stin-
gless bees. In Social Insects. An Evolutionary Approach to Castes
and Reproduction (W. Engels, Ed.), Springer-Verlag, Berlin. pp.
167–230.
Kerr, W.E., 1950. Genetic determination of castes in the genus
Melipona. Genetics 35: 143–152.
Kerr, W.E., R. Zucchi, J.T. Nakadaira and J.E. Botolo, 1962. Reproduc-
tion in the social bees (Hymenoptera: Apidae). J. N.Y. Entom. Soc.
70: 265–276.
Koedam, D., 1999. Production of queens, workers and males in the
stingless bee Melipona favosa (Apidae: Meliponinae): Pattern in
time and space. Neth. J. Zool. 49: 289–302.
Koedam, D., F.A.L. Contrera and V. L. Imperatriz-Fonseca, 1999.
Clustered male production by workers in the stingless bee Melipona
subnitida Ducke (Apidae, Meliponinae). Insectes Soc. 46: 387–
391.
Michener, C.D., 1974. The Social Behavior of the Bees. Harvard Uni-
versity Press, Cambridge, Mass. 404 pp.
Moo-Valle, H., J.J.G. Quezada-Euán and T. Wenseleers, 2001. The effect
of food reserves on the production of sexual offspring in the sting-
less bee Melipona beecheii (Apidae, Meliponini). Insectes Soc. 48:
398–403.
Paxton, R.J., L.R. Bego, H. Ruhnke, F.L.W. Ratnieks and J.G. Quezada-
Euan, 2001. Social evolution in stingless bees: are the workers or is
the queen in control of male production? Abstract XXVII Int. Ethol.
Conf. Tübingen. pp. 236–237.
Peters, J. M., D.C. Queller, V.L. Imperatriz-Fonseca, D.W. Roubik and
J.E. Strassmann, 1999. Mate number, kin selection and social con-
flicts in stingless bees and honeybees. in: Proc. R. Soc. London
Series B 266: 379–384.
Ratnieks, F.L.W., 2001. Heirs and spares: caste conflict and excess queen
production in Melipona bees. Behav. Ecol. Sociobiol. 5: 467–473.
Ratnieks, F.L.W. and H.K. Reeve, 1992. The structure of conflict in
monogyne hymenopteran societies: processes that reduce conflict in
advanced eusocial species. J. Theor. Biol. 158: 33– 65.
Ratnieks, F.L.W., 1988. Reproductive harmony via mutual policing by
workers in eusocial Hymenoptera. Am. Nat. 132: 217–236.
Sakagami, S.F., 1982. Stingless bees. In: Social Insects (H.R. Hermann,
Ed.). New York, Academic Press. pp. 361 –424.
Sommeijer, M.J. and H.H.W. Velthuis, 1977. Worker oviposition in or-
phan colonies of Melipona favosa. Proc. VIIIth Int. Congr. I.U.S.S.I.
Wageningen: 315 –316.
Sommeijer, M.J., F.T. Beuvens and H.J. Verbeek,1982. Distribution of
labour among workers of Melipona favosa : construction and pro-
visioning of brood cells. Insectes Soc. 29: 222–237.
Sommeijer, M.J. and L.L.M. de Bruijn, 1984. Social behavior of sting-
less bees: ‘bee-dances’ by workers of the royal court and the rhyth-
micity of brood cell provisioning and oviposition behavior. Behav-
iour 89: 229–315.
Sommeijer, M.J. and N.W.M. van Buren, 1992. Male production by lay-
ing workers in queenright colonies of Melipona favosa (Apidae,
Meliponinae). In: Biology and Evolution of Social Insects (J. Billen,
Ed.), Leuven University Press, Leuven. pp. 89–98.
Sommeijer M.J., T.X. Chinh and F.J.A.J. Meeuwsen, 1999. Behavioural
data on the production of males by workers in the stingless bee
Melipona favosa (Apidae, Meliponinae) Insectes Soc. 46: 92–93
Veen van, J.W., 2000. Cell provisioning and oviposition in Melipona
beecheii (Apidae, Meliponinae), with a note on caste determination.
Apidologie 31: 411– 419.
Veen van, J.W., H.G. Arce and M.J. Sommeijer, 1992. Brood production
in Melipona beecheii in relation to dry season foraging. In: Biology
and Evolution of Social Insects (J. Billen, Ed.), Leuven University
Press. Leuven. pp. 81 – 88.
Velthuis, H.H.W. and M.J. Sommeijer, 1990. A new model for the ge-
netic determination of caste in Melipona. Proc. Int. Congr. IUSSI,
Bangalore: 173.
Velthuis, H.H.W. and M.J. Sommeijer, 1991. Roles of morphogenetic
hormones in caste polymorphism in stingless bees. In: Morpho-
genetic Hormones of Arthropodes (A.P. Gupta, Ed.), Rutgers Univ.
Press. pp. 346–383.
Wenseleers, T., F.L.W. Ratnieks and J. Billen, 2001. Conflict over caste
fate in social insects: a tragedy of the commons examined. In: Con-
flict from Cell to Colony, Chapter 10. Ph.D. thesis, University of
Leuven, Belgium. pp. 174– 195.