Cuticular hydrocarbon profiles correlate with caste, sex, and polyethism in the stingless bee Melipona solani (Hymenoptera: Meliponini)

Abstract Information exchange and nestmate recognition among workers of highly social insects are tasks that usually involve cuticular hydrocarbons (CHCs). Many studies in stingless bees have shown that CHCs carry information about nest origin, age, caste, sex, reproductive status, and their function in the colony. Thus, in this work, we characterised the CHC composition of queens, gynes, drones, and workers of Melipona solani by gas chromatography coupled to mass spectrometry. Any possible age-related difference was investigated by analysing the CHC profile of workers of different ages. We found quantitative and qualitative differences in the CHC composition between castes (workers and queens), and among drones, gynes, and queens. Similarly, polyethism (age of workers) correlated with CHCs, which allowed the separation of three groups: (i) foragers/guards; (ii) nurses; and (iii) larvae, pupae, and recently emerged workers. We discuss the possible function of these compounds in the stingless bees’ recognition interactions.


Introduction
Cuticular hydrocarbons (CHCs) are the most abundant chemical compounds in insects' epicuticle and act as a barrier against water loss (Blomquist & Bagneres, 2010).However, several other functions have been assigned to insect hydrocarbons (Borges et al., 2012;Wyatt, 2003).Bees possess a unique CHC profile with which they carry cues about their nest origin, age, caste, sex, reproductive status, and function in the colony (Blomquist & Bagneres, 2010;Leonhardt et al., 2013;Nunes et al., 2009b).Recent studies have demonstrated that these compounds vary among colonies, castes, and individuals of different ages (Ferreira-Caliman et al., 2010;Nunes et al., 2010;Poiani et al., 2014).In addition, each individual has a cuticular chemical signature responsible for mutual communication that is essential for maintaining the integrity of the colony (Borges et al., 2012).
Stingless bees comprise the most diverse group of all eusocial bee species (Ascher & Pickering, 2020;Ayala et al., 2013;Michener, 2013); they occur in perennial colonies that range from several dozens to thousands of workers (Mel endez-Ramirez et al., 2013).To preserve the cohesion of the colony, stingless bees (Hymenoptera, Meliponini) have developed highly complex behavioural and physiological traits, like temporal polyethism and reproductive specialisation (Amano et al., 2000;Michener, 2013).Reproductive males only serve to provide a way to disperse genes and reduce the effect of inbreeding, unless they stay in their colony (Cameron et al., 2004).Virgin queens (also known as gynes) stay in the nest, leave during the swarm, supersede the dominant queen, or are killed, depending on what happens inside the colony (Imperatriz-Fonseca & Zucchi, 1995).The physogastric queen maintains the cohesion of the colony by influencing the behaviour and physiology of workers, which in turn are responsible for every task inside and outside of the colony (Gracioli-Vitti et al., 2004;Michener, 2013).
Division of labour among workers in a colony is related to individual age; they perform different activities sequentially, from activities needed inside the colony to activities that are required outside the colony: idle callows (recently emerged workers, 1day-old bees), workers at the comb area (nurses, 10day-old bees), and workers performing tasks outside (guards and foragers, >20-day-old bees) (Ferreira-Caliman et al., 2010;Nunes et al., 2010;Poiani et al., 2014;Sol-Balbuena et al., 2018).
The stingless bee Melipona solani is distributed in Southeast Mexico and Central America; this bee has cultural and economic importance due to its use in meliponiculture (Cortopassi-Laurino et al., 2006) and crop pollination (e.g., Persea americana and Bixa orellana; Ayala et al., 2013;Quezada-Eu an, 2018).Furthermore, due to its characteristics (medium-sized bee with the ability to buzz pollinate), this stingless bee is a potential effective buzz pollinator of Solanaceae (Mel endez-Ram ırez et al., 2018;Sol ıs-Montero et al., 2018).In previous work with M. solani, Alavez-Rosas et al. ( 2017) described the role of chemical cues during food recruitment; the authors found that the CHC composition is a qualitative variable among foragers from different colonies and among the different body parts.However, little is known about CHC chemistry among castes and different ages in a caste.Information about CHCs is highly valuable for understanding the role of these compounds in the chemical communication system of M. solani, and this information could be useful in the conservation and management of this stingless bee (Cortopassi-Laurino et al., 2006;Poiani et al., 2014;Quezada-Eu an, 2018).Thus, we used M. solani to investigate the variation of CHCs in a caste polyphenism (queens and workers) b) sex (males and females) c) polyethism (worker ages).

Insects
All specimens were obtained from four managed colonies of M. solani located in the Meliponary of El Colegio de la Frontera Sur in Tapachula Chiapas, Mexico.Colonies were previously inspected by an experienced beekeeper, who considered them in good health: free of fungi and parasites, with a good resin accumulation, a functional physogastric queen, approximately 2500 adults, 8-10 brood combs, honey reserves of about 1.5 L, and 10 pots full of pollen.

CHCs in M. solani castes
CHCs were obtained by standard techniques (see below) from four colonies of M. solani.We used 12 gynes (3 gynes from each colony) and 16 workers (4 foragers/guards from each colony).Finally, we used three physogastric queens that came from three of those colonies.

CHCs in M. solani sexes
CHCs were obtained from 20 males (5 males from each colony), and we used the same extracts from females that we used for the castes' analysis.

CHCs in workers of different ages
Brood combs with late pupal stages were incubated at 28 C in darkness until the emergence of imagos.Then, these bees were marked with colour-coded numbers to keep a record of their age and were immediately transferred to new nests (30 Â 30 Â 30 cm wood boxes).These new nests were placed outside the lab, to allow bees to have outside activities.Sixteen specimens of each of the following groups were subject to CHC analysis, 1-day post-emergence workers (recently emerged worker), 10-day post-emergence workers (nurse), and >20-day post-emergence workers (forager/guard).Additionally, 16 larvae and 16 pupae were subject to the same analysis; the ages of the larvae and pupae were not determined.

CHC extraction
Bees were sacrificed by freezing at À20 C immediately before analysis.The extracts were obtained by placing dead bees in a vial with 1 mL of pure n-hexane for 1 min (Nunes et al., 2009b).Then, the insects were removed from the glass vial and the solvent was allowed to evaporate under a gentle stream of dry N 2 flow.Finally, the cuticular substances were suspended in 100 mL of n-hexane and stored at À20 C until analysis.

CHC identification
Chemical analyses were carried out in the laboratory of Chemical Ecology at El Colegio de la Frontera Sur in Tapachula, Chiapas, Mexico.All compounds were identified using mass spectrometry (electron-impact ionisation [EI] and chemical ionisation [CI]) and infrared spectroscopy data.Some compounds were confirmed by comparing retention indexes and mass spectra with those of synthetic standards (Sigma-Aldrich, Toluca, Mexico).To identify the position of the double bond in the alkenes, a set of samples from one of the colonies was treated with dimethyl disulphide (DMDS) (Shibahara et al., 2008).
To identify the compounds, extracts were analysed with gas chromatography-mass spectrometry (GC-MS), using a Varian Star model CP-3800 GC (Palo Alto, CA, USA).A DB-5 fused silica capillary column (30 m Â 0.25 mm internal diameter [ID]) was temperature programmed from 50 C (held for 2 min) to 280 C at 15 C min À1 ; it was then held at 280 C for 10 min.The temperature of the injector was held at 250 C. The GC was coupled to a Varian Saturn 2200 mass spectrometer and an integrated data system.EI was carried out at 70 eV, 250 C. CI mass spectra were obtained using a GC-MS Triple-Quadrupole Mass Spectrometer TQ8040 with the CI ion volume in Q1MS mode.The same DB-5 capillary column and GC conditions were used as in EI experiments.Methane was used as the CI reagent gas.CI conditions were the following: ion source pressure 2 Torr; ionisation energy 70 eV; electron multiplier voltage 1500 V; emission current 200 mA; ionisation temperature 150 C; mass range m/z 50-600; and scan time 1 s.
Infrared spectra were obtained with a Spectra analysis GC-infrared spectrophotometer DiscovIR-GC (Marlborough, MA, USA).

Statistical analyses
All analyses were carried out using R software (R Core Team, 2019).The areas of the peaks in each sample were transformed into percentages (each compound was divided by the sum of the entire CHC area in the corresponding chromatogram).Percentages were analysed by principal component analysis (PCA), in which we added 0.01 to all percentages to avoid zeros in the analysis.To evaluate the degree of similarity of the groups formed in the PCA, the Euclidean distances among data were measured and compared through a one-way analysis of variance (ANOVA), including the interaction as a factor and the distance as the response variable.Means were compared with the Tukey test (a ¼ 0.05).
62.1% of the variance of the data (PC 1 ¼ 38.2% and PC 2 ¼ 23.9%).Based on the Euclidean distances (Table 2), the cuticular composition of workers was more similar to males than to queens/gynes.The lowest Euclidean distance was between queens and gynes.
In addition to caste (workers vs queens) and sex (males vs females), polyethism was also a factor that distinguished three groups: (a) workers (foragers/guards), (b) nurses, and (c) larvae/pupae/recently emerged workers (Figure 2).The alkenes 9-pentacosene, 9-heptacosene, 7-heptacosene, 7-hexacosene, 8-hexacosene, and 9-hexacosene were relevant to define the worker (forager/ guard) group.Nurses were characterised by n-tetracosane and n-hexacosane.Finally, the larvae/pupae/recently emerged worker group was closely related to 9-tricosene, n-heptacosane, 9-nonacosene, and 9-hentriacontene.The chemical similarity was confirmed with the Euclidean distances (Table 3).The PCA explained 68.4% of the total variance of the data: factor 1 described 50.2% of the variation, while factor 2 described 18.2% of the variance.When considering all data (caste, sex, and polyethism) in one PCA analysis, four groups could be defined (Supplementary Information 1): (a) queens, (b) gynes, (c) workers (foragers/guards), and (d) bees that live inside the colony (males, nurses, recently emerged workers, pupae, and larvae).The sum of the first two factors extracted by PCA, applied to all 26 compounds described by the GC-MS analysis, explained 48.8% of the total variation.

Discussion
We found quantitative and qualitative differences in CHCs among the castes, sexes, and polyethism of M. solani.These differences might have a role in colony cohesion (Abdalla et al., 2003;Ferreira-Caliman et al., 2010, 2012;Nunes et al., 2009a).These findings contribute to understanding the correlation of CHCs and functional roles in the genus Melipona (Ferreira-Caliman et al., 2010;Kerr et al., 2004).CHCs of stingless bees are a complex and highly diverse blend of self-produced and environmentally derived compounds (Leonhardt et al., 2013).We determined the environmentally and genetically derived CHCs in this species by comparing workers at different ages, includingfor the first time for a stingless bee speciesthe analysis of CHCs from larvae and pupae.
Regarding the CHCs, M. solani bees were classified into four groups: gynes, queens, males, and workers (foragers/guards).This classification is in agreement with those known for Melipona (Abdalla et al., 2003;Ferreira-Caliman et al., 2010;Kerr et al., 2004).However, we found that the chemical similarity (measured with Euclidean distances) of the CHC profiles among gynes, queens, males, and workers is different for M. solani compared with other species of this genus.For example, the CHC profiles of Melipona marginata males are more closely related to those of gynes than to workers (Ferreira-Caliman et al., 2013); however, the CHC profiles were similar among them in M. solani.Only the gynes have nheptadecane, n-nonadecane, and 9-eneicosane.It is possible that these hydrocarbons and 9-methylheptacosane (which is the only methylated hydrocarbon present in gynes) function as virgin-queen-specific compounds.This finding is in agreement with a previous report in which a selective 9-alkene hydromethylation is a feature that differentiates a physogastric queen from a virgin queen (Abdalla et al., 2003;Pianaro et al., 2007).Gynes have 9-tricosene and 9-pentacosene, while physogastric queens have 9-methyltricosane, 9-methylpentacosane, and 9methylheptacosane.Methyl-branched alkanes have been found in males, gynes, and queens of other stingless bee species (Ferreira-Caliman et al., 2013;Nunes et al., 2009a), but we only found these compounds in queens and gynes.Nunes et al. (2017) suggested that the stingless bee queen CHCs are interpreted as a signal that the queen is present; thus, methyl-branched alkanes may serve as distinctive compounds of M. solani queens.Despite these notable differences, we found that the gynes and queens have the most similar CHC profiles.This finding is supported by the fact that gynes become queens and acquire those methyl-branched alkanes after copulation (Abdalla et al., 2003;Nunes et al., 2009a).
M. solani males produced almost the same compounds as workers, and they did not produce compounds that gynes and queens did; moreover, 9triacontene was found only in males.These results are in agreement with previous reports that Melipona bicolor males have chemical cuticular profiles that are more similar to the profiles of workers than queens (Abdalla et al., 2003;Pianaro et al., 2009).Similarly to Kerr et al. (2004), we found that workers are more similar to males than to queens regarding the cuticular compounds.Although males have a different cuticular composition than queens and workers, the chemical similarity that males have with queens and workers is the same.This phenomenon could be useful in colony division to overcome the lack of males, one of the principal challenges in the maintenance of Melipona colonies (Quezada-Eu an et al., 2001).According to the acceptance threshold hypothesis, which posits that recognition may depend on the similarity of CHC profiles (Nascimento & Nascimento, 2012), it should be possible to exchange males among colonies to disperse genes and facilitate copulation.However, behavioural studies are needed to test whether behaviour could be modulated by the similarity of cuticular chemistry.The compounds 9-tetracosene, 7-hexacosene, 8-hexacosene, 9-hexacosene, and 9-octacosene were only found in workers and possibly confer a distinctive CHC profile for this caste, as described in other stingless bees (Nunes et al., 2009b;Pianaro et al., 2009).Due to the high variability in CHC profiles of Melipona species, our study contributes to understanding the relationship between CHCs and behavioural patterns observed within this genus (Kerr et al., 2004;Nascimento & Nascimento, 2012).
The age of M. solani workers was correctly classified according to their functional CHC profiles.Larvae, pupae, and newly emerged workers showed a distinct chemical profile (qualitatively) compared to nurses and foragers/guards.The compounds 9-tricosene, n-tricosane, 7/9-pentacosene, n-pentacosane, 7-heptacosene, 9-heptacosene, n-heptacosane, 9nonacosene, and 9-hentriacontene were found in younger bees.Older bees (nurses and foragers/ guards) had those compounds found in younger stages in addition to 9-tetracosene and n-tetracosane for nurses, and n-hexacosane, 7-hexacosene, 8-hexacosene, 9-hexacosene, and 9-octacosene for foragers/guards.This phenomenon could be a consequence of aging.Leonhardt et al. (2013) documented a relationship between phylogenetic and environmental factors on the composition of the cuticular chemistry of tropical stingless bees.In this scenario, the alkenes 7-hexacosene, 8-hexacosene, and 9-hexacosene could be environmentally derived compounds because they were found only in bees that performed a task outside the nest (forgers/ guards).These workers (foragers and guards) fly out of the nest to collect pollen and resins from a wide variety of plants from which they can acquire these compounds.Stingless bees incorporate compounds from the surrounding environment in their cuticular chemical profiles, an action that may increase their defensive properties (K€ amper et al., 2019).Although 9-octacosene was found in nurses and workers (forager/guards), this compound could be environmentally derived, transferred from workers (foragers/ guards) to nurses through trophallaxis (Camazine et al., 1998).The exchange of CHCs among individuals is crucial in most of the activities that permit the survival of the colony, like food collection, division of labour, reproduction, and defence (Leonhardt, 2017;Michener, 2013;Nunes et al., 2011).Because n-tetracosane and n-hexacosane are correlated to nurses, the task-specificity of this caste probably relies on these two alkanes.These findings contrast with the hypothesis that linear alkanes do not have a role in chemical recognition (Abdalla et al., 2003;Sol-Balbuena et al., 2018).We found a higher similarity in the cuticular chemistry between nurses and younger stages (recently emerged workers, larvae, and pupae) than between nurses and workers (forager/guards) and between workers (foragers/guards) and younger stages.These results contrast with those reported for M. marginata (Ferreira-Caliman et al., 2010) and Scaptotrigona postica (Poiani et al., 2014); those researchers reported greater similarity between nurses and workers.However, our finding correlates with the strong influence of the environment on the cuticular chemistry of some stingless bees (Leonhardt et al., 2013).

Conclusions
M. solani bees produce qualitatively and quantitatively different CHCs.Gynes, queens, males, and workers (foragers/guards) have distinctive CHC profiles.The highest similarity among cuticular profiles is between queens and gynes, while the lowest similarity is between workers (foragers/guards) and gynes, and between workers (foragers/guards) and queens.Finally, workers of different ages have dissimilar CHC profiles; however, we found greater similarity between nurses and younger stages (recently emerged workers, larvae, and pupae) than between nurses and workers (forager/guards) and between workers (foragers/guards) and younger stages.Future studies are needed to clarify the role of these compounds in the recognition process in this species.More studies are also needed to determine aging-dependent CHC changes in males, gynes, and queens, and to measure the possible exchange of CHCs between workers (e.g., nurses with recently emerged workers, or nurses with foragers/guards).

Figure 1 .
Figure 1.The plot of the two principal components of the cuticular hydrocarbon profiles of Melipona solani.G, gynes; M, males; Q, queens; W, workers (foragers/guards).
the NIST library (mass spectra and normal alkane retention index [RI]).b Chemical ionisation in mass spectrometry and infrared spectroscopy data.c Compared with a synthetic standard.

Figure 2 .
Figure 2. The plot of the two principal components of the cuticular hydrocarbon profiles of workers of Melipona solani at different ages.L, larvae; N, nurses; P, pupae; R, recently emerged worker; W, workers (foragers/guards).

Table 1 .
Cuticular hydrocarbons found in the stingless bee

Table 2 .
Euclidean distances between the groups formed in the principal component analysis of the stingless bee Melipona solani.

Table 3 .
Euclidean distances between the different ages of workers formed in the principal component analysis of the stingless bee Melipona solani.