Open access journal: http://periodicos.uefs.br/ojs/index.php/sociobiology ISSN: 0361-6525 DOI: 10.13102/sociobiology.v60i3.317-322Sociobiology 60(3): 317-322 (2013) Centris aenea (Hymenoptera, Apidae): a Ground-nesting Bee with High Pollination Efficiency in Malpighia emarginata DC (Malpighiaceae) GA Oliveira1, CML Aguiar1, M Silva2, M Gimenes1 Introduction The production of West Indian cherry or acerola (Mal� pighia emarginata DC) is increasing in Brazil and a signifi- cant part of the fruit production has been exported mainly to the United States, Japan, and countries of the European Com- munity. The fruit is not only consumed fresh but is also used as raw material for the food and pharmaceutical industries (Musser, 1995). Brazilian agro-business has had a growing interest in commercializing acerola but pollinator deficit, which results in low fruit set, is one of the factors hindering increases in productivity (Schlindwein et al., 2006; Guedes et al., 2011). Interactions between M. emarginata and its pollina- tors are based primarily on the availability and collection of floral oils and pollen (Oliveira & Schlindwein, 2009; Vilhena et al., 2012; see Freitas et al., 1999, and Siqueira et al., 2011) for details on floral morphology). In M. emarginata, flower Abstract This study aimed to assess the efficiency of Centris aenea Lepeletier as pollinator of Malpighia emarginata DC. The pollination efficiency was determined according to three criteria: 1) the pollen deposition rate of M. emarginata and other plant species on the thorax ventral region of females, 2) the rate of M. emarginata fruit set after a single visit, and 3) the pollen deposition rate on the stigma after one single visit. We found that 59% of the pollen grains deposited on the ventral region of C. aenea came from flowers of M. emarginata. The fruit set after a single visit by C. aenea (21%) was higher than that reported previously for Centris tarsata Smith. We argue that pollina- tion efficiency of C. aenea can be related to its body size, since that females of C. aenea can carry out larger amounts of pollen grains of M. emarginata than C. tarsata, as well females of C. aenea are able to touch a larger stigma area, resulting in higher fruit sets after a single visit. Our study suggests that only one visit by C. aenea ensures fruit set in M. emarginata. Sociobiology An international journal on social insects 1 - Universidade Estadual de Feira de Santana - UEFS, Feira de Santana, Bahia, Brazil 2 - Faculdade de Tecnologia e Ciências, Salvador, Bahia, Brazil Article History Edited by Kleber Del-Claro, UFU, Brazil Received 20 July 2013 Initial acceptance 09 August 2013 Final acceptance 20 August 2013 Key words Pollination Deficit, Crop pollinator, Bee-plant interaction, Oil-collecting bee, West Indian cherry Corresponding author Cândida Maria Lima Aguiar Univ. Estadual de Feira de Santana Av. Transnordestina, Novo Horizonte 44036-900, Feira de Santana, Bahia Brazil E-Mail: candida.aguiar@gmail.com pollination is performed only by specialist floral visitors and its pollinators are bees (Apidae) belonging to tribe Centridini (Freitas et al., 1999; Siqueira et al. 2011; Guedes et al., 2011) (see Freitas et al.,1999 for scheme and details on pollinator’s behavior). Some cavity-nesting Centris species have been sug- gested as promising candidates for pollination of M. emar� ginata (Buchmann, 2004; Oliveira & Schlindwein, 2009). The main advantage of these species lies in the easiness to obtain and relocate its nests because they undergo nidifica- tion in movable trap-nests (Jesus & Garófalo, 2000; Roubik & Villanueva-Gutiérez, 2009; Aguiar & Garófalo, 2004; Pina & Aguiar, 2011). On the other hand, other studies have found that ground-nesting Centridini species such as Centris aenea Lepeletier and Centris varia (Erichson), reach high abundan- ce in some regions and visit a large number of M. emarginata flowers (Siqueira et al., 2011; Vilhena et al., 2012), and are important for their pollination services to the crop. RESEARCH ARTICLE - BEES GA Oliveira, CML Aguiar, M Silva, M Gimenes - Pollination Efficiency of Centris aenea 318 Comparative studies on the efficiency of several polli- nators are needed to support the choice of the best species to be managed for increases in crop productivity. Although several cavity-nesting bees actually have favorable charac- teristics for management as crop pollinators, the role of each species in crop pollination has been scarcely studied. Stu- dies on the nesting-cavity bees in crop areas have often fo- cused on the bee high ability to occupy trap-nests (Oliveira & Schlindwein, 2009; Pina & Aguiar, 2011; Magalhães & Freitas, 2012) rather than on the abundance of bee species in crop flowers (e.g. Vilhena et al., 2012) or the efficiency of pollinators (Freitas et al., 1999; Freitas & Paxton, 1998; Freitas et al., 2002). Preliminary samples obtained in the study area indica- ted that C. aenea, a ground-nesting bee, was the most abun- dant visitor on M. emarginata flowers whereas the cavity- nesting species Centris analis (Fabricius) and Centris tarsata Smith were much less frequent (unpublished data). Despite its low frequency in flowers, these two cavity-nesting species use the orchard as their nesting site (Pina & Aguiar, 2011), and females of C. analis provision their brood cells mainly with pollen of M. emarginata (unpublished data). C. aenea has been registered in other areas as the most frequent po- tential pollinator of M. emarginata flowers (Siqueira et al., 2011; Guedes et al. 2011; Vilhena et al., 2012) but its pollina- tion efficiency has not yet been studied. For this reason, this study aimed to assess the efficiency of this species as pollina- tor of M. emarginata, and to support the choice of a potential species for the management to pollination of this crop. Materials and Methods Study Site The sample was collected at an acerola orchard (12°16’S, 38°58’W) located in the countryside of the mu- nicipality Feira de Santana, Bahia, Brazil. The climate is se- miarid with average annual precipitation of 802 mm and ave- rage annual temperature of 24ºC (CEI, 1994). The orchard contains approximately 300 plants of M. emarginata, grown under natural conditions with no irrigation or pesticides. Flower visitors The visitors of M. emarginata flowers were sampled using an entomological net once a month, from July 2010 to June 2011, from 0700 to 1700 hours, for 30 minutes every hour. In each sample, two collectors walked throughout the orchard collecting bees in flowering plants. We measured twenty-five females of the most abun- dant species (C. aenea). Their entire body length (between the median ocellus and the end of the abdomen) and interte- gular width (the distance between wing bases) were measu- red with a digital paquimeter. The Reproductive System Reproductive biology tests (Dafni et al., 2005) were performed in November 2010 and April 2011, when more than 50% of the plants were in bloom. Two-hundred and for- ty flowers in pre-anthesis were previously bagged with voil bag and received the following treatments: 1- hand self-polli- nation (autogamy); 2- spontaneous self-pollination; 3- hand cross-pollination (xenogamy); 4- control 1, marked flowers kept under natural conditions and available to flower visitors (natural pollination). Flowers were kept bagged for approxi- mately 15 days until fruit set, when bags were removed and the fruit set for each treatment was quantified. Pollination Efficiency of Centris aenea The pollination efficiency of C. aenea was determined according to three criteria. First, the pollen deposition rate of M. emarginata and other plant species on the thorax ventral region of female C. aenea collected while visiting acerola flowers. To reach this goal, all pollen grains from the females ventral region (n=12) were removed with glycerinated gela- tin. Next, the pollen grain-glycerinated gelatin mixture was dissolved in warm water and centrifuged at 2500 rotations per 10 min; the pollen grains were transferred to assay tu- bes containing acetic acid for at least 24 h for dehydration and then the material was processed using the acetolysis te- chnique (Erdtman, 1960). After this chemical treatment, the pollen grains were kept for at least one hour in a glycerin aqueous solution at 50% for rehydration. For the analysis of the pollen types deposited on fe- males ventral region, four slides were prepared using pollen sample obtained in each female. Slides were prepared with colorless gelatin, and one slide also for each sample was pre- pared with gelatin colored with safranin at 1%. The slides were examined under an optical microscope to identify and count the pollen of M. emarginata and other pollen types. At least 1,000 pollen grains per sample were counted. The second criterion determined the rate of M. emar� ginata fruit set after a single visit by C. aenea. Fifty-two flo- wers in pre-anthesis were bagged for the experiment; twenty- five of them were emasculated. On the next day, flowers were exposed to a single visit by C. aenea and were rebagged to prevent further visitations until fruit set. Two procedures were conducted in this experiment: in the first, 77 flowers were kept bagged, insect visits forbidden; in the second, 78 flowers were exposed to unrestricted bee visits (natural polli- nation, control 2). Fruit set was determined 15 days later for both experiments. The third criterion analyzed pollen deposition rate on the stigma. The use of this technique aimed to complement the fruit set rate experiment after a single visit by C. aenea and count the pollen grains of M. emarginata that females of C. aenea were able to deposit on the stigma of flowers after Sociobiology 60(3): 317-322 (2013) 319 one single visit. Fifty flowers in pre-anthesis were bagged and on the next day, 20 flowers were exposed to a single visit by C. aenea (10 of which were emasculated). After the first visit, the stigmas were removed from flowers and fixed on slides with glycerinated gelatin colored with safranin. Ten of the remaining 30 flowers were left bagged to prevent visits and 20 (10 of which were emasculated) were exposed to un- restricted bee visits from 0700 to 1700 h. After this period, the stigmas were also removed and fixed in slides with glyce- rinated gelatin colored with safranin. The deposited pollen grains were counted in the laboratory. The Kruskal-Wallis test was used to compare the number of pollen grains deposited on the stigma in each tre- atment. Differences in fruit set rates among treatments of the reproductive system of M. emarginata and after the single vi- sit on C. aenea were analyzed using the Chi-square test (χ²). Results Pollination Biology and Flower Visitors Results of the reproductive biology experiments show the occurrence of fruit set in treatments with hand cross-pollination (14.5%) and natural pollination (control 1) (24%); fruit sets did not occur in spontaneous or hand self- pollination experiments (Table I). No significant difference in fruit set rates between natural and hand cross-pollination (χ²=1.5905, p=0.207) were found. The number of fruits pro- duced by natural pollination was different from the number of fruits obtained by hand self-pollination or spontaneous self-pollination (χ²=14.524, p<0.001). The number of fruits obtained by hand cross-pollination also differed from hand self-pollination and spontaneous self-pollination (χ²=8.3258, p=0.004). Fruits produced by natural pollination (N=96 fruits, control 1) had 2.95 ± 0.2 seeds per fruit. Pollination efficiency of Centris aenea Of the pollen grains (n=8,496) deposited on the ven- tral region of C. aenea, 59% came from flowers of M. emar� ginata. Other pollen types, such as Solanum paniculatum, Poincianella microphylla, Stigmaphyllon sp. and Salvia sp. were also found in the samples and comprised approximately 40% of all grains. Table 1. Fruit set of West Indian cherry (Malpighia emarginata) in a semiarid area in Brazil. Table 2. Number and proportion of fruits set per Malpighia emar�Malpighia emar� ginata flower after a single visit by Centris aenea, in a West Indian cherry crop area in a semiarid area in Brazil. Fruit set after a single visit by C. aenea was 21% (Ta- ble II). There was no difference in the number of fruits produ- ced after a single visit by C. aenea on emasculated (16%) and non-emasculated flowers (26%) (χ²=0.76, p=0.38). Similarly, results of these experiments did not differ from experiments obtained from fruit sets in flowers open to visitation (n = 78, 30.7%) (Control 2) (χ²=1.46, p=0.22) (Table II). Bagged flo- wers did not produce fruits. After the single visit by C. aenea, pollen grains were deposited on stigmas of all the 20 flowers studied. The num- ber of grains deposited on stigmas of emasculated flowers varied from 6 to 51 (31.7±12.9), whereas it varied from 13 to 67 (33±17.8) on non-emasculated flowers, and no diffe- rence in deposition rates among treatments was observed (Kruskal-Wallis H=10.7, p=0.9296). The number of pollen grains deposited on the stigmas of flowers with unrestricted visits were between 2 and 371 in emasculated flowers and between 6 and 844 in non-emasculated flowers. No signifi- cant difference was observed between these two treatments (Kruskal-Wallis H=10.7, p=0.18). Discussion Results of the reproductive biology experiments show that the population of M. emarginata we studied did not produce fruits in self-pollination experiments, only in cross- pollination experiments and is considered self-incompatible Flowers of M. emarginata were visited only by bees belonging to tribe Centridini: Centris flavifrons (Fabricius), C. aenea Lepeletier, C. analis (Fabricius), C. fuscata Lepe- letier, C. trigonoides Lepeletier, C. tarsata Smith, C. obsole� ta Lepeletier, C. sponsa Smith, Epicharis flava (Friese) and three other non-identified species of Centris. Among these, the most abundant species was C. aenea (66% of all bees), with females measuring 15.3±0.9 mm in length and 0.59± 0.23 mm in width. Trea tment Number of flowers Number of fruits set Flowers tha t set fruits (%) Sponta neous self-pollina tion 53 0 0 Ha nd self-pollina tion 53 0 0 Ha nd cross-pollina tion 55 8 14.5 Na tura l pollina tion (control 1) 54 13 24 Trea tment Number of flowers Number of fruits set Flower tha t set fruits (%) Flowers non-ema scula ted (exposed to one visit) 27 7 26 Flowers ema sculed (exposed to one visit) 25 4 16 Flowers ba gged (without visits) 77 0 0 Na tura l pollina tion (control 2) 78 24 30.7 GA Oliveira, CML Aguiar, M Silva, M Gimenes - Pollination Efficiency of Centris aenea 320 and dependent on biotic agents (Centridini bees) for pollina- tion. However, the fruit set rate by cross-pollination was low in unrestricted visits experiments (natural pollination), as also observed in other Brazilian regions (Freitas et al., 1999; Guedes et al., 2011). The lack or low fruit set rates for the crop by self-pollination was also observed by other resear- chers (Schlindwein et al., 2006; Guedes et al., 2011, Siqueira et al., 2011). Self-pollination rates between 11% and 35% were observed for several varieties in irrigated M. emargina� ta orchards (Siqueira et al., 2011). As such, low fruit set per inflorescence and dependence on entomophilus pollination seem to be traits of this plant species (Freitas et al., 1999; Siqueira et al., 2011). A plausible hypothesis to explain the low fruit set ob- served in M. emarginata is that this reproductive factor is directly related to the frequent ovule abortion in this plant species (Miyashita et al., 1964; Freitas et al., 1999), whe- reby few fruits are produced in spite of flower abundance. On the other hand, the low M. emarginata fruit set rate has also been explained by climatic conditions such as tempe- rature, rainfall and luminosity, by genetic factors, and the low abundance of orchard pollinators (Yamane & Nakaso- ne, 1961; Schlindwein et al., 2006, Carpentieri-Pípolo et al., 2008; Siqueira et al., 2011; Guedes et al., 2011). Guedes et al. (2011) recently registered pollination deficit in M. emargina� ta in a Brazilian semiarid region during the dry season, when pollinator abundance was much lower than during the rainy season. Those authors obtained a significantly higher fruit set after applying complementary hand cross-pollination, after natural pollination performed by bees during the dry season. In our study, pollination experiments were conducted whe- never pollinator abundance was high, and therefore the low fruit set rate after natural pollination might not be related to the lack of pollinators. Centris aenea was the most important of all species of Centris visiting flowers of M. emarginata, not only due to its abundance but also for its high floral fidelity, noticeable primarily by the frequent presence (more than 50% of all pol- len grains) on the thorax ventral region of M. emarginata. C. aenea was also the most abundant species on acerola flowers at our study site, particularly during blooming peaks (unpu- blished data). An even higher abundance (above 80%) of C. aenea on acerola flowers was reported in a study conducted in the semiarid region of Bahia State (Siqueira et al., 2011). These data show that C. aenea prefers foraging for pollen on M. emarginata flower in acerola orchards, thus a positive assessment of the bee efficiency as effective crop pollinator. The average number of pollen grains observed on the stigma after the single visit by C. aenea was enough to ferti- lize the ovules and produce fruits because only three pollen grains are required for fruit formation in M. emarginata. Cru- den (2000) argued that the number of pollen grains deposited on the stigma indicates pollinator efficiency and depends on several factors such as size of the stigma area and duration of stigmatic receptivity. Stigmatic receptivity in M. emarginata lasts more than six hours (Freitas et al., 1999; Schlindwein et al., 2006), a trait that can increase the exposure time of flowers to pollinators as well as the chances for the effective pollination to occur because a larger number of co-specific pollen grains can be deposited on the stigmas, as already known (Dafni et al., 2005). As a consequence, there would be fewer chances for stigma contamination with hetero-spe- cific pollen, avoiding clogging and leading to higher fruit sets and more effective pollinator visitations (Slaa & Biesmeijer, 2005). We observed that M. emarginata fruit set after a single visit by C. aenea (21%) was much higher than that found for C. tarsata (7%) (Freitas et al., 1999). They reported that more than one visit per flower by C. tarsata is necessary for a high fruit set to be achieved, whereas our study suggests that only one visit by C. aenea ensures fruit set in M. emarginata. The different efficiencies in pollination after a single visit between the two bee species might not be due to behavior because both species have the same behavior pattern during floral oil collection on these flowers (Freitas et al., 1999; Schlindwein et al., 2006; Siqueira et al., 2011). However, pollination efficiency can be related to body size because C. aenea is approximately 1.5 times bigger than C. tarsata (n= 9, length = 10.4 ± 0.5 mm; width = 3.5 ± 0.25 mm; personal observation). Body size can be related to the amount of pol- len remaining on the pollinator body and by consequence, the amount of pollen to be deposited on the stigma, as also observed by Cruden and Miller-Ward (1981) and by Cruden (2000). Due to its size, females of C. aenea can be more effi- cient in transporting larger amounts of pollen grains of M. emarginata than C. tarsata, as well females of C. aenea are able to touch a larger stigma area, resulting in higher fruit sets after a single visit. Pollination of the M. emarginata population studied was conducted primarily by a ground-nesting bee, C. aenea. This bee was a highly efficient pollinator very abundant in flowers, whereas other species of Centridini, including the cavity-nesting species were much less frequent. Maintaining areas with adequate soil conditions for nesting of C. aenea, and also surrounding areas with vegetation that supplements the supply of nectar and even pollen when M. emarginata blooming is low are recommended strategies to keep the lo- cal populations of this pollinator. Acknowledgments We thank to the National Council for Scientific and Technological Development, Brazil (CNPq) (Proc. #475715/2008-0; #562518/2010-0) and to the Fundação de Amparo à Pesquisa do Estado da Bahia (FAPESB) (TO APP0042/2009) for financial support for this project. C.M.L. Aguiar received a research fellowship from CNPq (Proc. 306403/2012-9), G.A. Oliveira received a M.Sc. scholarship Sociobiology 60(3): 317-322 (2013) 321 from CAPES and M. Silva received a postdoctoral fellowship from FAPESB. We are grateful to the Universidade Estadual de Feira de Santana (UEFS) for logistic support. References Aguiar, C.M.L. & Garófalo, C.A. (2004). Nesting biology of Centris (Hemisiella) tarsata Smith (Hymenoptera: Apidae: Centridini). Rev. Bras. Zool., 21: 477-486. Buchmann, S.L. (2004). Aspects of Centridini biology (Cen� tris spp.) importance for pollination, and use of Xylocopa spp. as greenhouse pollinators of tomatoes and other crops. In B.M. Freitas & J. O. Pereira (Eds.), Solitary bees: con- servation, rearing and management for pollination (pp. 203- 211). Fortaleza: Universidade Federal do Ceará. Carpentieri-Pípolo, V., Neves, C.S.V.J., Bruel, D.C., Souza, S.G.H. & Garbúglio, D.D. (2008). Frutifi cação e desenvol-2008). Frutificação e desenvol- vimento de frutos de aceroleira no norte do Paraná. Cienc. Rural, 38: 1871-1876. CEI - Centro de Estatística e Informações da Bahia. (1994). Informações Básicas dos Municípios Baianos: Região Pa- raguaçu. Salvador: Centro de Estatística e Informações da Bahia. Cruden, R.W. & Miller-Ward, S. (1981). Pollen-ovule ratio, pollen size, and the ratio of stigmatic area to the pollen- bearing area of the pollinator: an hypothesis. Evolution, 35: 964-974. Cruden, R.W. (2000) Pollen Grains: Why So Many? Plant. Syst. Evol., 222: 143-165. Dafni, A., Kevan, P.G. & Husband, B.C. (2005). Practical pollination biology. Cambridge: Enviroquest. Erdtman, G. (1960). The acetolysis method. A revised des- cription. Svensk Botanisk Tidskrift, 54: 561-564. Freitas, B.M. & Paxton, R.J. (1998). A comparison of two pollinators: the introduced Honey bee Apis mellifera and an indigenous bee Centris tarsata on cashew Anacardium occi� dentale in its native range of NE Brazil. J. Appl. Ecol., 35: 109-121. Freitas, B.M., Alves, J.E., Brandão, G.F. & Araújo, Z.B. (1999). Pollination Requirements of West Indian cherry (Malpighia emarginata) and its putative pollinators, Centris bees, in NE Brazil. J. Agric. Sci., 133: 303-311. Freitas, B.M., Paxton, R.J. & Holanda-Neto, J.P. (2002). Identifying pollinators among an array or flower visitors, and the case of inadequate cashew pollination in NE Brazil. In P. Kevan & V. L. Imperatriz-Fonseca (Eds.), Pollinating bees: the conservation link between agriculture and nature (pp. 229-224). Brasília: Ministry of Environment. Guedes, R.S., Zanella, F.C.V., Martins, C.F. & Schlindwein, C. (2011). Déficit de polinização da aceroleira no período seco no semiárido paraibano. Rev. Bras. Frutic., 33: 465- 471. Jesus, B.M.V. & Garófalo, C.A. (2000). Nesting behaviour of Centris (Heterocentris) analis (Fabricius) in southeastern Brazil (Hymenoptera, Apidae, Centridini). Apidologie, 31: 503-515. Magalhães, C.B. & Freitas, B. M. (2012). Introducing nests of the oil-collecting bee Centris analis (Hymenoptera: Api- dae: Centridini) for pollination of acerola (Malpighia emargi� nata) increases yield. Apidologie, 44: 234-239. doi: 10.1007/ s13592-012-0175-4 Miyashita, R.K., Nakasone, H.Y. & Lamoureux, C.H. (1964). Reproductive morphology of acerola (Malpighia gla� bra). Hawaii Agr. Exp. Stat., Technical Bulletin, Honolulu, Hawaii, n. 63. Musser, R.S. (1995). Situação atual e perspectivas da acero-Situação atual e perspectivas da acero- la. In A.R. São José & R. E. Alves (Eds.), Acerola no Brasil: Produção e mercado (pp. 4-6). Vitória da Conquista: Univer- sidade Estadual do Sudoeste da Bahia. Oliveira, R. & Schlindwein, C. (2009). Searching for a mana- geable pollinator for acerola orchards: the solitary oil-collec- ting bee Centris analis (Hymenoptera: Apidae: Centridini). J. Econ. Entomol., 102: 265-273. Pina, W.C. & Aguiar, C.M.L. (2011). Trap-nesting bees (Hy- menoptera: Apidae) in orchards of acerola (Malpighia emar� ginata DC) in a semiarid region in Brazil. Sociobiology, 58, 379-392. Roubik, D.W. & Villanueva-Gutiérrez, R. (2009). Invasive Africanized honey bee impact on native solitary bees: a pol- len resource and trap analysis. Biol. J. Linn. Soc., 98: 152- 160. Schlindwein, C., Martins, C.F., Zanella, F.C.V., Alves, M.V., Carvalho, A.T., Darrault., R.O., Duarte, Jr. J.A., Oliveira, M.D., Ferreira, A.G., Guedes, R.S., Ferreira, R.P., Pinto, C.E., Silveira, M.S. & Vital, M.T.A.B. (2006). Diagnóstico e manejo dos polinizadores de mangabeira e aceroleira. In Anais do VII Encontro sobre Abelhas (pp. 443-454). Ribei- rão Preto: Universidade de São Paulo. Siqueira, K.M.M.; Martins, C.F.; Kiill, L.H.P. & Silva, L.T. (2011). Estudo comparativo da polinização em variedades de aceroleiras (Malpighia emarginata DC, Malpighiaceae). Rev. Caatinga, 24: 18-25. Slaa, J. & Biesmeijer, K. (2005). Flower constancy. In A. Da- fni, P.G. Kevan, B.C. Husband (Eds.), Practical pollination biology (pp. 381-400). Cambridge: Enviroquest. Vilhena, A.M.G.F., Rabelo, L.S., Bastos, E.M.A.F. & Augus- to, S.C. (2012). Acerola pollinators in the savanna of Central Brazil: temporal variations in oil-collecting bee richness and GA Oliveira, CML Aguiar, M Silva, M Gimenes - Pollination Efficiency of Centris aenea 322 a mutualistic network. Apidologie, 4: 51–62. doi: 10.1007/ s13592-011-0081-1 Yamane, G. & Nakasone, H.H. (1961). Pollination and fruit set studies of acerola Malpighia glabra L. in Hawai. Proc. Am. Soc. Hortic. Sci., 78: 141-148.