Open access journal: http://periodicos.uefs.br/ojs/index.php/sociobiology ISSN: 0361-6525 DOI: 10.13102/sociobiology.v60i3.283-288Sociobiology 60(3): 283-288 (2013) Food Niche Overlap Among Neotropical Carpenter Bees (Hymenoptera: Apidae: Xylocopini) in an Agricultural System DM Carvalho, CML Aguiar, GMM Santos Introduction Most studies on bee niches have focused on the diversity of floral resources used and the partitioning of these resour- ces (e.g., Wilms et al., 1996; Goulson & Darvill, 2004; Aguiar, 2003; Aguiar et al., 2013; Andena et al., 2012; Figueiredo et al., 2013). However, guilds of flower visitors can also be studied from other perspectives, such as activity time (Carvalho et al., 2010; Santos et al., 2013). The partitioning of temporal niches may facilitate the coexistence of ecologically similar species (Kronfeld-Schor & Dayan, 2003). The level of complementarity or redundancy in the niches of flower-visiting insects is an important mechanism underlying the relationship between biodiversity and ecosys- tem functioning (Blüthgen et al., 2006; Hoehn et al., 2008; Santos et al., 2013). However, little empirical evidence exists about the level of complementarity among pollinators (Hoehn et al., 2008). Different species may have complementary niches and perform different functional roles, (Santos et al., 2010), thereby presenting different realized niches (Begon Abstract In the present study, we used niche overlap analysis and a network approach to inves- tigate the use of floral resources by carpenter bees (Xylocopa spp.). We assessed the frequency of visit to different plant species and the activity time of carpenter bees in an agricultural system. Six species of carpenter bees were collected visiting flowers. Among the 48 interactions that were theoretically possible in the interaction network, only 19 were recorded (Connectance = 39.58%). The temporal overlap between pairs of species measured by the Shannon index (0 to 0.648) was lower than dietary overlap (0 to 0.967). The network analysis also showed that bees separated their niches more strongly in the temporal dimension (E = 0.72, P < 0.001) than in the dietary dimension (E = 0.55, P < 0.001). The levels of dietary and temporal overlap were strongly corre- lated with each other, as well as the time of highest frequency of visit coincided with the time of availability of resources by the most important plants (Moringa oleifera, Passiflora edulis, and Solanum palinacanthum). The correlation between dietary and temporal overlap is biologically explained by the presence of plants that structure the system by exerting a strong influence not only on the plant choice by foraging bees, but also on the time of resource collection. Sociobiology An international journal on social insects Universidade Estadual de Feira de Santana - Feira de Santana, Bahia, Brazil Article History Edited by Kleber Del-Claro, UFU, Brazil Received 19 June 2013 Initial acceptance 30 July 2013 Final acceptance 07 August 2013 Key words Carpenter bees, bee-plant network, trophic niches, temporal activity Corresponding author Gilberto M. M. Santos Departamento de Ciências Biológicas Univ. Estadual de Feira de Santana Feira de Santana, Bahia, Brazil 44036-900 E-Mail gmms.uefs@gmail.com et al., 2006). Niche complementarity requires some degree of trophic or temporal specialization, whereas niche redun- dancy is associated with high niche overlap and functional redundancy (Blüthgen & Klein, 2011). However, two species may present redundancy on a dimension (e.g., diet) and com- plementarity in another dimension (e.g., temporal patterns of resource use) of their trophic niches (Santos et al., 2013). In the present study, we used niche overlap analysis and a network approach to investigate the use of floral re- sources by carpenter bees (Xylocopa spp.). We assessed the frequency of visit to different plant species and the activity time of bees in an agricultural system. These bees are attrac- tive candidates for crop pollination, because they have broad diet, long activity season, and are active under low illumina- tion levels (Keasar, 2010). Indeed, several crops, such as pas- sion fruits, tomato, and melons, can be favored by the polli- nation service rendered by carpenter bees (Hogendoorn et al., 2000; Sadeh et al., 2007; Benevides et al., 2009; Yamamoto et al., 2010; Keasar, 2010; Benevides et al., in press). RESEARCh ARTICLE - BEES DM Carvalho, CML Aguiar, GMM Santos - Food Niche Overlap Among Neotropical Carpenter Bees 284 Plant Species Carpenter bee species X. frontalis (X01) X. grisescens (X02) X. nigrocincta (X03) X. ordinaria (X04) Xylocopa sp.1 (X05) X. suspecta (X06) Anacardium occidentale L. (P01) - 1 - - - - Bixa orellana L. (P02) 5 1 - - - - Cucurbita pepo L. (P03) - 1 - - - - Moringa oleifera Lam. (P04) 23 3 4 - 1 7 Passiflora edulis Sims (P05) 51 - 2 1 - 4 Portulaca oleracea L. (P06) 1 - - - - - Solanum palinacanthum Dunal (P07) 3 - 1 1 - 2 Solanum stipulaceum Willd. ex Roem.& Schult. (P08) - - - 1 - - Material and Methods Study area We collected data on bee–plant interactions in the municipality of Feira de Santana, Bahia, northeastern Brazil (12º15’25”S, 38º57’54”W). The climate of the region is semi- arid, with an average annual temperature of 23.5 °C and an average annual rainfall of 867 mm (MMA 2009). The study was carried out at the Bocaiúvas Farm, that covers an area of 23.4 ha, where crops of fruits and vegetables dominate the landscape, including Citrus sinensis L (orange), Citrus limon- um L. (lime), Mangifera indica L. (mango), Psidium guajava L. (guava), Passiflora edulis Sims (passion fruit), Capsicum annuum L. (papikra), Cucurbita sp. (pumpkin), and Cucumis anguria L. (bur cucumber). All plants are cultivated with no chemical additives. Sampling We used as a sampling unit each record of bee-plant interaction. We recorded flower-visiting bees, plants visited, visit time, temperature, relative air humidity, and luminosity each hour during each sampling. In order to characterize the guild of carpenter bees that forages on flowers in an agricul- tural system, we carried out monthly sampling from August 2011 to July 2012. Before the first monthly sampling, we Data analysis We used the importance index (Ij) to evaluate the im- portance of each plant species for each bee species, as well as the importance of each bee species to each plant species. This index varies from 0 to 1, and it tends to 1 when a plant species (or a bee species) has many interactions in the com- munity or a large number of exclusive interactions (Murray, 2000). We measured the overlap in both trophic niche dimen- sions (dietary and temporal) for each pair of bee species with the Schoener index (1986). This index (NOih) varies from 0 to 1, and we considered the level of overlap low when the va- lue found was lower than 0.3, moderate when the value was higher than 0.3 and lower than 0.5, and high when the value was higher than 0.5. Following Santos et al. (2013), we used a network ap- proach to determine the degree of specialization in the diet and temporal activity of flower-visiting bees with the E index (Araújo et al. 2008). We organized the data on bee species and activity time as adjacency matrices and represented them as bipartite graphs using the package bipartite 2.00 for R (Dormann et al., 2009). We used linear regressions and Spearman correlations to test for relationships between temperature, air humidity, bee richness, bee abundance, and degree of niche overlap in pairs of bee species. Table 1. Carpenter bee species and plants visited for floral resources in an agricultural system. selected a pathway crossing places rich in flowering plants. In each sampling session, two collectors captured carpenter bees visiting flowers, using entomological nets, along the same transect 1 km long. The sampling effort per specimen of plant in blossom was 5 min. We carried out sampling from 0600 to 1800, summing up 12 samples of 12 h, or 244 h of sampling per collector. All flowering plant species were sam- pled each hour, as collectors walked the entire transect once every hour. Results We collected six species of carpenter bees (113 indivi- duals) visiting flowers of eight plant species (Table 1). In the interaction network we recorded 19 out of 48 theoretically possible bee-plant interactions (Connectance = 39.58%). The low connectance of the bee-plant network is related to single- tons, as four plant species were visited only by a single bee species, and one bee species visited only one plant species (Fig. 1). Moringa oleifera (P04), Passiflora edulis (P05), and Sociobiology 60(3): 283-288 (2013) 285 Solanum palinacanthum (P07) stood out by their number of interactions with bees. These species received 91% of the bee visits. Xylocopa frontalis (X01) was the most abundant bee (N = 83, 73%) and visited the highest richness of plant spe- cies (Tables 1 and 2) Carpenter bees visited flowers throughout the day, but with higher frequency of visits at three times (Fig. 2). The first peak of individuals visiting flowers occurred between 0600 and 0800 (32% of the individuals), the second between 1100 and 1300 (41%), and the third at 1500 (17%). We compared niche overlap (dietary and temporal) for six Xylocopa species (Table 2). The values calculated for species with few individuals should be interpreted with caution, since they may not reflect the real local overlap in the use of the resources. The dietary overlap varied from 0 to 0.967. Out of 15 pairs of species analyzed, three pairs showed low overlap Table 2. Richness of visited plants (Spla), abundance of individuals in each bee species (Nbees), and trophic niche overlap (NOih, Schoener in- dex) between pairs of species of carpenter bees flower-visiting in an agricultural system. Temporal niche overlap dimension (upper triangle) and dietary niche overlap dimension (lower triangle). Spla Nbees X. frontalis X. grisescens X. nigrocincta X. ordinaria Xylocopa sp1 X. suspecta Xylocopa frontalis (Olivier, 1789) 5 83 - 0.108 0.611 0.325 0.145 0.568 Xylocopa grisescens Lepeletier, 1841 4 5 0.337 - 0.143 0.000 0.000 0.385 Xylocopa nigrocincta Smith, 1854 3 7 0.599 0.500 - 0.286 0.143 0.648 Xylocopa ordinaria Smith, 1874 3 3 0.369 0.000 0.429 - 0.333 0.308 Xylocopa sp.1 1 1 0.277 0.500 0.571 0.000 - 0.154 Xylocopa suspecta Moure & Camargo, 1988 3 13 0.621 0.500 0.967 0.462 0.538 - 0,00 0,25 0,50 P06 (01) P01 (01) P03 (01) P02 (06) P08 (01) P07 (07) P05 (58) P04 (38) 0,00 0,25 0,50 0,75 X05 (01) X03 (07) X06 (13) X04 (03) X02 (06) X01 (83) Fig. 1 - Importance index (Ij; Murray) of the plant species that provide floral resources for carpenter bees (Ij = 0.002 to 0.361) and of the flower-visiting bees (Ij = 0.004 to 0.624). Bee abundance is provided in parenthesis. Plants and Bees codes on table 1. DM Carvalho, CML Aguiar, GMM Santos - Food Niche Overlap Among Neotropical Carpenter Bees 286 (NOih < 30%), seven pairs showed intermediate overlap (30% < NOih < 50%), and five pairs showed high overlap (NOih > 50%). Temporal overlap was lower than dietary overlap, va- rying from 0 to 0.648. Out of 15 pairs of species analyzed in terms of temporal overlap, eight showed low overlap (NOih < 30%), four pairs showed intermediate overlap (30% < NOih < 50%), and only three pairs showed high overlap (NOih > 50%; Table 2). The highest value of overlap in the dietary niche (0.967) was found between X. nigrocincta and X. suspecta, which resulted primarily from the concentration of visits both in the flowers of M. oleifera and P. edulis. strongly in the temporal dimension (E = 0.72, P < 0.001) than in the dietary dimension (E = 0.55, P < 0.001; Fig. 4). Discussion Among Xylocopa bees, X. frontalis was the species with the highest number of interactions and with the broadest trophic niche. This species is known for its generalistic fora- ging behavior and for exploring a relatively broad diversity of floral resources (Silva, 2009; Yamamoto et al., 2012). However, an analysis of the pollen collected by female X. frontalis, X. grisescens, and other carpenter bee species re- vealed that, in spite of being generalistic, carpenter bees tend to explore more intensively a subset of the floral resources available, which contributes to the narrowing of their dietary niche and increases dietary overlap between species (Silva, 2009). The three peaks of carpenter bee foraging on flowers observed in the present study and the high overlap in bee Fig. 2 – Number of individuals of carpenter bees (Xylocopa spp) captured visiting flowers in an agricultural system at each sampling time. 0 5 10 15 20 25 30 06:00 07:00 08:00 09:00 10:00 11:00 12:00 13:00 14:00 15:00 16:00 N um be r of in di vi du al s time other X. suspecta X. frontalis Fig. 3 – Correlation between dietary and temporal overlap among carpenter bees (Xylocopa spp) (y = 0.063 + 0.4815*x; r = 0.5557; p = 0.0315) in an agricultural system. There was no significant relationship between bee abundance, bee species richness, plant species richness, temperature, and relative air humidity. However, we found a significant positive relationship between the level of tem- poral overlap and the level of dietary overlap (y = 0.063 + 0.4815*x; r = 0.56; p = 0.03) (Fig. 3). The network analy- sis showed similar results; bees separated their niches more Fig. 4 – Graphs showing a larger weighted clustering coefficient in the bee-plant network than in the bee-time network. Dark gray pen- tagons represent carpenter bees (Xylocopa spp), light gray circles represent in graph A the plants visited, and in graph B the time of visit. Lines represent links and their thickness is proportional to the number of observations. Plants and Bee codes on table 1. A B Sociobiology 60(3): 283-288 (2013) 287 activity at these times seem to be related to the temporal avai- lability of the floral resources of three important plants. In the morning, the foraging activity of bees was more intensive at early hours, a period that coincides with the opening of M. oleifera flowers, which occurs from 0300 on (Jyothi et al., 1990) and of S. palinacanthum flowers, whose anthesis begins around 0600 (Carvalho et al., 2001). The second peak was recorded around midday, which coincides with the an- thesis and availability of the resources of P. edulis flowers at 1200 (Benevides et al., 2009; Siqueira et al., 2009). In the mid afternoon, there was an increase in the frequency of car- penter bees on flowers, which is consistent with the observa- tions made by Carvalho et al. (2001), who studied the visitors of S. palinacanthum. The availability of floral resources by P. edulis from midday on strongly influenced the overlap of carpenter bee niches, since this plant received the highest number of visi- tors. As in our study, Benevides et al. (2009) recorded the highest frequency of bee visits to P. edulis flowers between 12:30 and 13:30 in passion fruit crops in southeastern Brazil, in which X. frontalis and X. ordinaria were the most frequent carpenter bees. On the other hand, Varassin et al. (2010) found a constant visit rate of X. frontalis to P. edulis flowers between 14:00 and 18:00, and concluded that the visit rate of this carpenter bee was not influenced by anthesis timing or nectar volume or concentration. At times of low visit frequency of Xylocopa to the monitored plants, we presume that the females changed their foraging routes to explore other plants in the surroundings. Likewise, Jyothi et al. (1990) observed that the temporal ac- tivity of Xylocopa on M. oleifera depended not only on the availability of nectar provided by this plant, but also on the availability of resources produced by other plants in the sur- roundings, which could be more attractive to foraging fema- les at some times of the day. As pointed out by Silva (2009) and Yamamoto et al. (2012), in agricultural systems, Xylocopa species interact with both cultivated and wild plants. In the present study, abiotic factors (temperature and relative humidity) did not affect the visit frequency of car- penter bees or the overlap in dietary and temporal niches. Although there are several examples of a relationship between the foraging activity of bees and abiotic factors, there are also examples of the daily activity of bees explained by the temporal availability of floral resources. According to Stone et al. (1999), the time of resource production by plants re- presents a window of opportunity for foraging bees and may also be a driver of bee activity. In the studied system, the levels of dietary and tem- poral overlap were strongly correlated with each other, as well as the time of highest visit frequency of bees coincided with the time of availability of resources by most important plants. The correlation between dietary and temporal overlap is biologically explained by the presence of plants that struc- ture the system by exerting a strong influence not only on the plant choice by foraging bees (bee diets), but also on the time of resource collection. Acknowledgments Several colleagues contributed to the present study in different ways: J.J. Resende helped us in fieldwork, Maise Silva identified bee species, Efigenia Melo identified plant species and Marco A.R. Mello gave invaluable suggestions to an early version of the manuscript. The Brazilian Council for Scientific and Technological Development (CNPq) and the Bahia Research Foundation (FAPESB) funded the pre- sent study. CNPq granted G.M.M. Santos and C.M.L. Aguiar research productivity fellowships. References Aguiar, C.M.L. (2003). Utilização de recursos florais por abelhas (Hymenoptera: Apoidea) em uma área de Caatinga (Itatim, Bahia, Brasil). Rev. Bras. Zool., 20: 457-467. Aguiar, C.M.L., Santos, G.M.M., Martins, C.F. & Presley, S.J. (2013). Trophic niche breadth and niche overlap in a guild of flower-visiting bees in a Brazilian dry forest. Apido-Apido- logie, 44: 153-162. Andena, S. R., Santos, E.F. & Noll, F.B. (2012). Taxonom-Taxonom- ic diversity, niche width and similarity in the use of plant resources by bees (Hymenoptera: Anthophila) in a cerrado area. J. Nat. Hist., 46: 1663-1687. Araújo, M.S., Guimarães Jr., P.R., Svanbäck, R., Pinheiro, A., Guimarães, P., Reis, S.F. & Bolnick, D.I. (2008). Ne- twork analysis reveals contrasting effects of intraspecific competition on individual vs. population diets. Ecology, 89: 1981-1993. Begon, M., Townsend, C.R. & Harper, J.L. (2006). Ecology: from individuals to ecosystems. Oxford: Blackwell. Benevides, C.R., Gaglianone, M.C. & Hoffmann, M. (2009). Visitantes florais do maracujá-amarelo (Passiflora edulis f. flavicarpa Deg. Passifloraceae) em áreas de cultivo com dife- rentes proximidades a fragmentos florestais na região Norte Fluminense, RJ. Rev. Bras. Entomol., 53: 415-421. Benevides, C.R., Evans, D.M. & Gaglianone, M.C. (2013). Pollinators of Passifloraceae and the structure of their ne- twork in a fragmented lowland Atlantic forest. Sociobiology, 60: 297-307. doi: 10.13102/sociobiology.v60i3.295-305. Blüthgen, N., Menzel, F. & Blüthgen, N. (2006). Measuring specialization in species interaction networks. BMC Ecol., 6: 1-12. Blüthgen, N. & Klein, A.M. (2011). Functional complemen- tarity and specialisation: why biodiversity is important in plant-pollinator interactions. Basic Appl. Ecol., 12: 282-291. Carvalho, A.M.C. & Oliveira, P.E.A.M. (2010). Estrutura da DM Carvalho, CML Aguiar, GMM Santos - Food Niche Overlap Among Neotropical Carpenter Bees 288 guilda de abelhas visitantes de Matayba guianensis Aubl. (Sapindaceae) em vegetação do cerrado. Oecologia Australis, 14: 40-66. doi:10.4257/oeco.2010.1401.02. Carvalho, C.A.L. de, Marques, O.M., Vidal, C.A. & Neves, A.M.S. (2001). Comportamento forrageiro de abelhas (Hy- menoptera, Apoidea) em flores de Solanum palinacanthum Dunal (Solanaceae). Rev. Bras. Zool., 3: 35-44. Dormann, C.F., Fründ, J., Blüthgen, N. & Gruber, B. (2009). Indices, graphs and null models: analyzing bipartite ecologi- cal networks. Open Ecol. J., 2: 7-24. Figueiredo, N.; Gimenes, M.; Miranda, M.D. & Oliveira- Rebouças, P. (2013). Xylocopa Bees in Tropical Coastal Sand Dunes: Use of Resources and Their Floral Syndromes. Neotrop. Entomol., 42: 252-257. doi 10.1007/s13744-013- 0121-9. Goulson, D. & Darvill, B. (2004). Niche overlap and diet breadth in bumblebees; are rare species more specialized in their choice of flowers? Apidologie, 35: 55-63. Hoehn, P., Tscharntke, T., Tylianakis, J.M. & Steffan-Dewen- ter, I. (2008). Functional group diversity of bee pollinators increases crop yield. Proc. R. Soc. B, 275: 2283-2291. doi: 10.1098/rspb.2008.0405. Hogendoorn, K., Steen, Z. & Schwarz, M.P. (2000). Native Australian carpenter bees as a potential alternative to intro- ducing bumble bees for tomato pollination in greenhouses. J. Apicult. Res, 39: 67-74. Jyothi, P.V., Atluri, J.B. & Reddi, C.S. (1990). Pollination ecology of Moringa oleifera (Moringaceae). Proc. Indian Acad. Sci. (Plant Sci.). 100: 33-42. Keasar, T. (2010). Large Carpenter Bees as Agricultu- ral Pollinators. Psyche, 2010: 1-7, Article ID 927463, doi:10.1155/2010/927463. Kronfeld-Schor, N. & Dayan, T. (2003). Partitioning of time as an ecological resource. Annu. Rev. Ecol. Evol. Syst., 34: 153-181. MMA - Ministério do Meio Ambiente. (2009). Secretaria Es- tadual de Meio Ambiente Programa Nacional de capacitação de gestores ambientais – PNC. Gestão Ambiental Comparti- lhada – GAC. Plano Municipal de Meio Ambiente de Feira de Santana. Brasil. Feira de Santana. Murray, K.G. (2000). The importance of different bird spe- cies as seed dispersers. In: N.M. Nadkarni & N.T. Wheelwri- ght (Eds.), Monteverde: ecology and conservation of a tropi- cal cloud forest (pp. 245-302), Oxford University Press. Sadeh, A., Shmida, A. & Keasar, T. (2007). The carpenter bee Xylocopa pubescens as an agricultural pollinator in green- houses. Apidologie, 38: 508-517. Santos, G.M.M., Aguiar, C.M.L. & Mello, M.A.R. (2010). Flower-visiting guild associated with the Caatinga flora: trophic interaction networks formed by social bees and social wasps with plants. Apidologie, 41: 466-475, doi:10.1051/ apido/2009081. Santos, G.M.M., Carvalho, C.A.L., Aguiar, C.M.L., Macêdo, L.S.S. & Melo, M.A.R. (2013). Overlap in trophic and tem-Overlap in trophic and tem- poral niches in the flower-visiting bee guild (Hymenoptera, Apoidea) of a tropical dry Forest. Apidologie, 44: 64-74, doi:10.1007/s13592-012-0155-8. Schoener, T.W. (1986). Resource partitioning. In: J. Kikkawa & D.J. Anderson (Eds.) Community ecology - pattern and process (pp. 91-126). London: Blackwell Scientific Silva C.I. (2009). Distribuição espaço-temporal de recursos florais utilizados por espécies de Xylocopa (Hymenoptera, Apidae) e interação com plantas do cerrado sentido restrito no triângulo mineiro. Tese de Doutorado, Universidade Fe- deral de Uberlândia, Brasil. Siqueira K.M.M., Kiill, L.H.P., Martins C.F., Lemos, I.B., Monteiro, S.P., Feitoza, E.A. (2009). Ecologia da polinização do maracujá-amarelo, na região do Vale do Submédio São Francisco. Rev. Bras. Frutic., 31: 1-12. Stone, G.N., Gilbert, F., Willmer, P., Potts, S., Semida, F., Zalat, S. (1999). Windows of opportunity and temporal struc- turing of foraging activity in a desert solitary bee. Ecol. En- tomol., 24: 208-221. Varassin, I.G., Ximenes, B.M.S., Moreira, P.A., Zanon, M.M.F., Elbl, P., Löwenberg-Neto, P., & Melo, G.A.R. (2012). Produção de néctar e visitas por abelhas em duas es-Produção de néctar e visitas por abelhas em duas es- pécies cultivadas de Passiflora L. (Passifloraceae). Acta Bot. Brasilica, 26: 251-255. Yamamoto, M., Silva, C.I., Augusto, S.C., Barbosa, A.A.A. & Oliveira, P.E. (2012). The role of bee diversity in pollina-The role of bee diversity in pollina- tion and fruit set of yellow passion fruit (Passiflora edulis forma flavicarpa, Passifloraceae) crop in Central Brazil. Api-Api- dologie, 43: 515-526, doi: 10.1007/s13592-012-0120-6.