Sociality in the phylogenetically basal allodapine bee genus Macrogalea (Apidae: Xylocopinae): implications for social evolution in the tribe Allodapini

Recent phylogenetic studies based on DNA sequence data indicate that the tropical African bee genus Macrogalea is the sister group to the remaining extant allodapine fauna, whereas previously it was thought to be a distal group. This leads to some fundamental changes in our understanding of social evolution in the allodapines. Earlier studies suggested that Macrogalea showed only weak forms of social behaviour and these were not well characterized. However, large samples of Macrogalea zanzibarica presented here show that this species exhibits marked social behaviour. Nearly half of nests collected contained two or more adult females, with up to 10 females per nest. Brood are reared progressively and brood ages within colonies are staggered, giving rise to colonies with very mixed age structures and therefore frequent opportunities for alloparental care. Ovarian dissections indicate non‐simple forms of reproductive partitioning within colonies and most multi‐female colonies show evidence that more than one female has contributed to egg production, though reproductive episodes among colony members are frequently asynchronous. Some females show signs of much higher wing wear than their nestmates, but always show signs of previous reproduction. Reproductive division of labour appears to be temporally marked, ovarian differentiation among nestmates is linked to relative body size, but permanent worker‐like or queen‐like castes appear to be absent. This is similar to the communal, continuously brooded and multivoltine behaviour of some tropical halictine species and may be due to the aseasonal nature of brood development in tropical regions. Patterns of per capita brood production indicate large benefits to multi‐female nest occupancy, and sex allocation is strongly female biased. These findings strongly suggest that the capacity for complex social interactions and alloparental care is an ancestral trait for all of the extant allodapine lineages. Therefore comparisons among extant allodapines are unlikely to throw light on the initial origin of social behaviour, though they may uncover origins of true caste behaviour and reversals to solitary nesting. Sex ratios in Macrogalea and most other allodapine genera, spanning a broad phylogenetic and ecological range, suggest that one or more allodapine traits have provided persistent selection for female‐biased sex allocation.     

A single origin of large colony size in allodapine bees suggests a threshold event among 50 million years of evolutionary tinkering

Evolutionary origins of highly eusocial organization involving morphological castes have been very rare, yet these origins have often led to enormous diversification and ecological success. This suggests that once an apparently severe selective barrier to highly eusocial behaviour is overcome, major new adaptive landscapes open up. One would therefore expect a discontinuity in patterns of evolutionary change across this barrier. However, we do not know if highly eusocial organization has evolved incrementally from less complex societies, or if it has involved some kind of evolutionary leap. Our study examines this issue using colony size data from 33 allodapine bee species, with a crown age of ca. 47 Mya. Our species cover all major allodapine clades, and include Exoneurella tridentata, the only known allodapine with morphologically discrete castes. Phylogenetic analyses indicate a strong effect of phylogeny on the evolution of maximum brood size, but the effect of phylogeny on maximum colony size (number of adults) depends on whether E. tridentata is excluded or included in analyses. We found no evidence of punctuational change in maximum colony or brood sizes over the phylogeny as a whole, but colony and brood sizes in E. tridentata fall well beyond variation among the other allodapines. Colony size in E. tridentata therefore represents an evolutionary outcome that does not fit within the kinds of incremental changes found in other allodapines. We propose that E. tridentata indicates the crossing of an important threshold, and this has entailed some very unusual ecological circumstances.     

Brood insurance via protogyny: a source of female-biased sex allocation

Sex allocation patterns and colony productivity are examined in Exoneura nigrescens, a social allodapine bee. As for previous studies on Australian allodapine bees, numerical sex ratios were strongly female biased in the smallest broods, but neared equality in larger broods. Local fitness enhancement has been suggested previously to explain female-biased allocation in allodapine bees. Here, we propose an alternative model, the 'insurance model', which predicts protogyny and, as a consequence, female-biased sex allocation in small broods with decreasing bias in larger broods. Because allodapine broods are reared progressively in an open burrow, broods require that adult females be present throughout their development in order to survive to maturity. If mothers invest in daughters (alloparents) first, these daughters can rear younger, dependent brood in cases in which orphaning occurs. If such daughters behave as surrogate mothers, then investment in them by mothers should not be regarded as investment in female sex allocation per se, giving rise to apparently female-biased broods. The model predicts a pattern of sex ratios as a function of total brood size that very closely match empirical data from E. nigrescens.     

FEMALE-BIASED SEX RATIOS IN A FACULTATIVELY SOCIAL BEE AND THEIR IMPLICATIONS FOR SOCIAL EVOLUTION

Montane populations of the Australian allodapine bee, Exoneura bicolor, are characterized by high levels of cooperative nesting and strongly female-biased sex ratios. A conspecific population from heathland shows much lower levels of cooperative nesting and lower levels of female bias. In both habitats, sex-ratio bias is greatest in the smallest brood sizes and becomes successively less biased in larger broods. Parity is approached in the largest heathland colonies, but not for any brood-size category in montane areas. Adult intracolony relatedness is moderately high for colonies in both reused and newly founded nests in the montane habitat, but probably low or zero for newly founded nests in heathland. Colony efficiency, measured as the number of brood per adult, increases with colony size in both habitats, suggesting that cooperation between females increases mean female fitness. It is argued that patterns of sex allocation are consistent with nonlinear fitness-return models, in which the mean reproductive value of daughters increases with the number of daughters produced in a brood. Such increases probably arise from a number of social interactions, including cooperative brood defense, increased task efficiency, and lower per capita costs in nest construction. The term “local fitness enhancement” is introduced here to describe these effects collectively. The female-biased ratios should lower selective thresholds for sib-directed altruism, at least in the earlier stages of colony development. It is argued that local fitness enhancement facilitates eusociality in allodapine bees and could also play a role in other haplodiploid taxa, provided cooperative nesting largely involves sisters, colony efficiency increases with colony size, and optimal colony sizes are only achieved after two or more generations after founding.     

Reproductive hierarchies in the African allodapine bee Allodapula dichroa (Apidae: Xylocopinae) and ancestral forms of sociality

The social organization of allodapine bees has been described in detail for most genera, although there remains a notable gap for one major lineage, the genus Allodapula . Here, we provide the first detailed study of social organization in Allodapula dichroa . Colony sizes are small and the frequency of cooperative nesting is low compared with other allodapine taxa, but there is very clear evidence for reproductive differentiation among adult nestmates. Reproductively dominant females tend to be larger than their nestmates and have much higher levels of wing wear, suggesting that they perform most foraging activities. Multi-female colonies have: (1) lower rates of complete brood absence, suggesting a substantial benefit to cooperative nesting; and (2) larger numbers of brood, suggesting that the presence of a second adult female leads to a greater reproductive output. These data suggest a major phylogenetic split in the form of social organization within the allodapines. In the genus Macrogalea (sister clade to all other allodapines), body size does not preclude young females from laying eggs, and there appears to be, at most, weak reproductive queues. However, in most other allodapines, reproductive hierarchies are prominent and younger and/or smaller females queue for reproductive opportunities, adopt permanently subordinate roles, or disperse. Interestingly, the most common forms of reproductive hierarchies in allodapines do not involve subordinates undertaking foraging roles before reproduction, but instead involve the delaying of both reproduction and foraging. This has implications for the understanding of suggested developmental ground plans in the early stages of social evolution.  © 2009 The Linnean Society of London, Biological Journal of the Linnean Society , 2009, 97 , 520–530.     

The biology of Brevineura froggatti and phylogenetic conservatism in Australian allodapine bees (Apidae,Allodapini)

Summary: We present the first data on the social biology of the allodapine bee, Brevineura froggatti. Colony sizes are small, and only 12.5% of nests contained more than two females. Brood rearing occurs throughout the year, including winter, as it does in the other species of Brevineura studied to date. In both Brevineura species, per capita brood production is much higher in multi-female nests than single female nests, raising the question of why more colonies aren't multifemale. The occurrence of small colony sizes, despite large benefits to group living, differs strongly from species of the sister clade Exoneura. These findings, combined with previous allodapine studies, indicate conservatism in voltinism, brood phenology and colony size within, but not between, Australian allodapine genera.     

Sociality in the Australian Allodapine Bee Brevineura elongata: Small Colony Sizes Despite Large Benefits to Group Living

Allodapine bees (family Apidae, subfamily Xylocopinae) provide substantial material for investigating the evolution of sociality because of their wide variation in colony size, life history traits, and caste differentiation. Two recent studies have shown that the Australian allodapine genus Brevineura exhibits a strong increase in per capita brood production (PCBP) in social colonies compared to single-female nests. However both species previously examined, B. xanthoclypeata and B. froggatti, show relatively few multi-female nests, which is puzzling considering the apparently large advantages for group-living in these species. Here we show that in a third species, B. elongata, there are also substantial benefits for group living, involving increased PCBP and a greatly reduced likelihood of nests without brood. As expected from these observations, we also found strongly female biased sex allocation. Nevertheless only a small percentage of nests contained more than one adult female, similar to the other two Brevineura species, raising the question of why multifemale colonies are not more common in this genus. Solving this puzzle will throw light on conditions that constrain sociality when group living apparently provides major advantages.     

Molecular phylogenetics of allodapine bees,with implications for the evolution of sociality and progressive rearing

Allodapine bees have long been regarded as providing useful material for examining the origins of social behavior. Previous researchers have assumed that sociality arose within the Allodapini and have linked the evolution of sociality to a transition from mass provisioning to progressive provisioning of brood. Early phylogenetic studies of allodapines were based on morphological and life-history data, but critical aspects of these studies relied on small character sets, where the polarity and coding of characters is problematic. We used nucleotide sequence data from one nuclear and two mitochondrial gene fragments to examine phylogenetic structure among nine allodapine genera. Our data set comprised 1506 nucleotide positions, of which 402 were parsimony informative. Maximum parsimony, log determinant, and maximum likelihood analyses produced highly similar phylogenetic topologies, and all analyses indicated that the tropical African genus Macrogalea was the sister group to all other allodapines. This finding conflicts with that of previous studies, in which Compsomelissa + Halterapis formed the most basal group. Changing the basal node of the Allodapini has major consequences for understanding evolution in this tribe. Our results cast doubt on the previous hypotheses that progressive provisioning and castelike social behavior evolved among lineages leading to the extant allodapine taxa. Instead, our results suggest that mass provisioning in Halterapis is a derived feature and that social behavior is an ancestral trait for all allodapine lineages. The forms of social behavior present in extant allodapines are likely to have resulted from a long evolutionary history, which may help explain the complexity of social traits found in many allodapine bees.     

Behavioural specialization in pre-reproductive colonies of the allodapine beeExoneura bicolor (Hymenoptera: Anthophoridae)

Summary Exoneura bicolor is a univoltine allodapine bee common in montane forests of southern Australia, where it exhibits a semisocial/quasisocial colony organization. Within-nest behaviour in postemergence autumn nests ofExoneura bicolor was recorded with the aim of studying behavioural specialization in pre-reproductive colonies. Ten complete colonies were transferred to purpose-built observation nests shortly before brood eclosion in late summer. Behaviour within observation nests was recorded for periods of up to 44 days after establishment, covering a period when colonies are preparing for overwintering. Dispersal of females and brood rearing do not occur at this time, although some females may become inseminated. Analyses of data using multivariate techniques indicated four distinguishable behavioural castes, designated here as Guards, Nest Absenters, Nest Modifiers and Non-recruits. This represents a higher degree of behavioural specialization than recorded to date for other allodapines. Behaviours performed by Guards and Nest Absenters are likely to involve considerable risks, but benefit the colony as a whole, so that some nestmates in prereproductive colonies exhibit altruism that frequently aids adult siblings or cousins. The males in our study were fed by females via trophallaxis and two of the males participated in nest maintenance tasks. Our results suggest that autumn colonies ofE. bicolor form well-integrated behavioural units even though brood rearing does not commence until the following spring.     

Sociality and sex allocation in an Australian allodapine bee Braunsapis protuberans

Abstract  Herein the first detailed study of sociality and life history in an Australian species of the allodapine bee genus Braunsapis is provided. Colonies of B. protuberans were sampled from the Great Sandy National Park in Queensland from late winter until later summer. Single-female colonies appear to have suffered substantially higher rates of total brood loss, or failure to produce brood, than multifemale colonies, but rearing efficiency in colonies with at least some brood did not differ between these two colony types. There was marked reproductive skew in multifemale colonies and ovary size was strongly linked to relative body size. However, females with smaller body size and smaller ovaries also had lower levels of wing wear, suggesting that they do not specialise in foraging. These patterns are similar to an African species, Braunsapis vitrea , and suggest that colonies may principally comprise a dominant reproductive female who provides food for her juvenile offspring, along with a variable number of subordinate females who may be waiting to inherit their natal nest in the event of their mother's death. However, it is likely that in such events, the remaining female(s) will care for orphaned brood as some single-female colonies contained substantial older brood, even though the sole female had little or no wing wear. Such alloparental care in the case of orphaning may help explain the high level of female bias in sex ratios.     

Sex Ratios in a Socially Parasitic Bee and Implications for Host-Parasite Interactions

Obligate social parasites of Hymenoptera, known as inquilines, have received enormous attention due to the elaborate adaptations they exhibit for exploiting their hosts, and because they have frequently been used to infer sympatric speciation. Their population biology can be difficult to infer as they are both rare and difficult to extract from host nests. Sex allocation has been studied for very few inquilines of social Hymenoptera. Here we report sex ratio patterns in the allodapine bee Inquilina schwarzi, which is an obligate social parasite of another allodapine, Exoneura robusta. We show that the sex ratio of this inquiline varies with its brood number, it is female-biased in the smallest broods, but becomes more even in larger broods, where the population-wide sex ratio is close to parity. We argue that this pattern of bias is consistent with local resource competition, where inquiline females compete to inherit their natal colony. We also argue that extremely female-biased sex ratios of the host species, combined with overall sex ratio parity in the parasite, may help ameliorate disparity in effective population sizes between these two species which are locked in an evolutionary arms race.     

Strong constraints to independent nesting in a facultatively social bee: quantifying the effects of enemies-at-the-nest

Constraints to independent nesting play a key role in the understanding of social evolution in insects, but the source and the magnitude of such constraints are poorly known for many key taxa. In allodapine bees it is known that solitary nesting females have low rates of successful brood rearing and that this drives selection for cooperative nesting. It has been hypothesized that these constraints are due to the presence of enemies-at-the-nest, such as ants, but no direct link has been demonstrated between such enemies and colony failure. We set up an experiment in which solitary founded nests of an Australian allodapine bee, Exoneura nigrescens, were either protected from non-flying predators or left unprotected, and compared the resulting colony survival and brood production rates. We found that protected colonies have much higher rates of survival and that the constraints to independent nesting are extreme, with a mean of less than one offspring per nest at the end of the brood rearing period. This means that cooperative nesting is essential for this species to persist in its habitat. Received 6 July 2007; revised 5 November 2007; accepted 12 November 2007.     

Sociality in a Malagasy allodapine bee, Macrogalea antanosy, and the impacts of the facultative social parasite, Macrogalea maizina

Social parasitism has been researched extensively in many taxa of social insects, including ants, wasps and bees. However, little research has been done on allodapine bees, a taxon that has numerous independent origins of social parasitism. This study looks at two species of Macrogalea from Madagascar, one of which was previously believed to be a social parasite. Macrogalea is an important genus to study as it is the sister clade to all other allodapine genera, and the species of Macrogalea in Madagascar diverged recently, meaning that the study of a social parasite in this genera would provide insights into the very early stages of social parasite evolution. Macrogalea maizina was determined to be facultatively parasitic based on the presence of many traits that are common to other allodapine social parasites. The host, Macrogalea antanosy, was found to be quasisocial, with most females within a colony being able to reproduce. This has unique consequences for a parasitic strategy, as any invading parasite has no need to remove a queen or suppress the reproduction of the other colony members, a strategy that has been commonly observed for facultative parasites in other taxa. Received 10 May 2005; revised 22 July 2005; accepted 24 August 2005.     

Life cycle and social behavior in a heathland population of the allodapine bee,Exoneura bicolor (Hymenoptera: Apidae)

Summary Previous studies of the allodapine beeExoneura bicolor have utilised a predominantly social population from tall montane forests. In this report we examine aspects of the sociality and biology of a population from subcoastal heathland in western Victoria, Australia. Entire colonies, nesting inXanthorrhoea minor, were collected at monthly intervals over 12 months between February 1989 and January 1990. In addition, three collections fromMelaleuca squarrosa were made over three consecutive winter months. The heathland population differs from the montane population in several aspects associated with sociality: (i) it shows lower levels of co-occupancy in both newly founded and re-used nests; (ii) foundress dispersal occurs over a larger proportion of the year; and (iii) there may be a partical second brood produced in late summer, during and after adult eclosion of the first brood. The occurrence of a partial second brood gives rise to opportunities for sib-rearing and eusociality. Colony sizes tend to be larger inM. squarrosa thanX. minor, but still smaller than colonies from the montane population. Egg-laying commences several weeks earlier in multifemale colonies than in single female colonies, before spring dispersal begins. It appears that multifemale colonies are more successful at preventing brood loss than single females and that the mean number of brood per female is larger in multifemale colonies than in single female colonies. These two populations, which display differences in several life history traits, provide an ideal system for studying the mechanisms maintaining social behavior.     

Sex ratios, local fitness enhancement and eusociality in the allodapine bee Exoneura richardsoni

Previous studies of a facultatively eusocial allodapine bee, Exoneura richardsoni Rayment, indicated that high levels of cooperative nesting among close relatives seem to be maintained by benefits that lead to increases in per capita brood production. These traits could lead to local fitness enhancement, which in turn could select for female-biased sex ratios. We show here that sex investment ratios in this species are female-biased in small colony sizes, becoming progressively male-biased in larger colonies, consistent with expectations for local fitness enhancement, but not explainable by alternative models. Our results support previous suggestions that local fitness enhancement can lead to sex ratio bias in primitively social Hymenoptera, but differ from previous studies by suggesting that patterns of bias could lower selective thresholds for sib-directed altruism in small colonies, but have an opposing effect in large colonies.     

Irreversible transfer of brood care duties and insights into the burden of caregiving in incipient subterranean termite colonies

1. In incipient termite colonies, biparental brood care rapidly shifts towards alloparental brood care. This transition was suggested to recapitulate the evolutionary trajectory from subsocial wood roach ancestors to eusociality in termites. 2. Incipient colonies of the subterranean termite Coptotermes gestroi (Wasmann) were investigated to determine if this transition was irreversible and if the burden of brood care on the first alloparents could be measured. To do so, the initial size of the work force necessary for an incipient colony to survive once the brood care became alloparental was determined. 3. The results of the study show that within 5 months after foundation, brood care duties were fully transferred to workers and the primary reproductives became irreversibly dependent on these workers for survival, reproduction, and colony growth. 4. Once the brood care became strictly alloparental, the presence of a single worker was enough to maintain the survival of the king and queen, confirming that ‘reversed parental care’ was also achieved. However, major brood loss and suppressed egg‐laying activity from the queen was observed, suggesting that the burden of brood care was too high for a single worker to absorb. Therefore, once brood care has shifted to alloparents, a critical number of workers is necessary to prevent brood loss and initiate colony growth. 5. As the initial cost of brood care is rapidly absorbed with colony growth in termites, the performance in brood care of the first few alloparents in a subsocial wood roach ancestor may have contributed to the emergence of eusociality in this clade.     

Egg‐size plasticity in Apis mellifera: Honey bee queens alter egg size in response to both genetic and environmental factors

Social evolution has led to distinct life‐history patterns in social insects, but many colony‐level and individual traits, such as egg size, are not sufficiently understood. Thus, a series of experiments was performed to study the effects of genotypes, colony size and colony nutrition on variation in egg size produced by honey bee (Apis mellifera) queens. Queens from different genetic stocks produced significantly different egg sizes under similar environmental conditions, indicating standing genetic variation for egg size that allows for adaptive evolutionary change. Further investigations revealed that eggs produced by queens in large colonies were consistently smaller than eggs produced in small colonies, and queens dynamically adjusted egg size in relation to colony size. Similarly, queens increased egg size in response to food deprivation. These results could not be solely explained by different numbers of eggs produced in the different circumstances but instead seem to reflect an active adjustment of resource allocation by the queen in response to colony conditions. As a result, larger eggs experienced higher subsequent survival than smaller eggs, suggesting that honey bee queens might increase egg size under unfavourable conditions to enhance brood survival and to minimize costly brood care of eggs that fail to successfully develop, and thus conserve energy at the colony level. The extensive plasticity and genetic variation of egg size in honey bees has important implications for understanding life‐history evolution in a social context and implies this neglected life‐history stage in honey bees may have trans‐generational effects.     

Social polymorphism in the Australian small carpenter bee, Ceratina (Neoceratina) australensis

The bee tribe Ceratinini provides important insights into the early stages of sociality. Many arthropods exhibit behaviours precursory to social life, including adult longevity, parental care, nest loyalty and mutual tolerance, yet the origins of social behaviour are few. Here we describe the biology of a facultatively social bee, Ceratina (Neoceratina) australensis, which exhibits all of the preadaptations for successful group living, yet presents ecological and behavioural characteristics that seemingly disfavour frequent colony formation. This species is socially polymorphic with both solitary and social nests collected in sympatry. Social colonies consist of two foundresses, one contributing both foraging and reproductive effort and the second which remains at the nest as a passive guard. Cooperative nesting provides no overt reproductive benefits over solitary nesting in this population, although brood survival tends to be greater in social colonies. Maternal longevity, subsociality and bivoltine nesting phenology in this species favour colony formation, while dispersal habits and offspring longevity may inhibit more frequent social nesting in this and other ceratinines.     

The Influence of Maternal Quality on Brood Sex Allocation in the Small Carpenter Bee,Ceratina calcarata

In the twig‐nesting carpenter bee, Ceratina calcarata, body size is an important component of maternal quality, smaller mothers producing significantly fewer and smaller offspring than larger mothers. As mothers precisely control the sex and size of each offspring, smaller mothers might compensate by preferentially allocating their investment towards sons. We investigated whether variation in maternal quality leads to variation in sex allocation patterns. At the population level, the numerical sex ratio was 57% male‐biased (1.31 M/F), but the investment between the sexes was balanced (1.02 M/F), because females are 38% larger than males (1.28 F/M). Maternal body size explained both sex allocation pattern and size variation among offspring: larger mothers invested more in individual progeny and produced more female offspring than smaller mothers. Maternal investment in offspring of both sexes decreased throughout the season, probably as a result of increasing maternal wear and age. The exception to this pattern was the curious production of dwarf females in the first two brood cell positions. We suggest that the sex ratio distribution reflects the maternal body size distribution and a constraint on small mothers to produce small broods. This leads to male‐biased allocation by small females, to which large mothers respond by biasing their allocation towards daughters.     

Comparison of social and solitary nesting carpenter bees in sympatry reveals no advantage to social nesting

Facultatively social bees allow for comparisons of social and solitary behaviour under similar environmental conditions. When such bees are polymorphic within the same population it provides a special and somewhat unusual opportunity to examine factors leading to cooperative (social) behaviours where many parameters such as environment are not variable. Some species of bees in the genus Xylocopa offers such opportunities. Studies of these bees often point to guarding and resource limitation as factors leading to social nesting. The large carpenter bee Xylocopa virginica L. (Hymenoptera: Apidae) is one species that exhibits both solitary and social nesting behaviour within the same populations. This paper first compares social and solitary nests in an Ontario population of X. virginica to determine if there is a reproductive advantage to social nesting. Following this, a series of possible explanations for social nesting, and the roles of females in social nests are examined. Social nests have similar brood sizes to solitary nests and productivity as a function of colony size is reduced with increasing number of foundresses. Additional foundresses are not effective guards, do not prevent parasitism, do not likely perform significant work, and do not assume foraging with the loss of a primary foundress. © 2014 The Linnean Society of London, Biological Journal of the Linnean Society, 2014, 113 , 998–1010.