The importance of plant diversity in maintaining the pollinator bee , Eulaema nigrita ( Hymenoptera : Apidae ) in sweet passion fruit fields

The euglossine bee Eulaema nigrita plays an important role for the pollination of native and economically important plants, such as the sweet passion-fruit Passiflora alata. E. nigrita uniquely collects the nectar from the flowers of P. alata, nevertheless, it needs to visit other plants to collect pollen, nectar and other resources for its survival. There are two methods to identify the species of plants used by bees in their diet: by direct observation of the bees in the flowers, and through identification of pollen grains present in brood cells, feces, or in the bees’ body. In order to identify the other plants that E. nigrita visits, we analyzed samples of pollen grains removed from the bee’s body in the course of the flowering period of P. alata. Among our results, the flora visited by E. nigrita comprised 40 species from 32 genera and 19 families, some of them used as a pollen source or just nectar. In spite of being a polyletic species, E. nigrita exhibited preference for some plant species with poricidal anthers. P. alata which has high sugar concentration nectar was the main source of nectar for this bee in the studied area. Nonetheless, the pollinic analysis indicated that others nectariferous plant species are necessary to keep the populations of E. nigrita. Studies such as this one are important since they indicate supplementary pollen-nectar sources which must be used for the conservation of the populations of E. nigrita in crops neighbouring areas. In the absence of pollinators, growers are forced to pay for hand pollination, which increases production costs; keeping pollinators in cultivated areas is still more feasible to ensure sweet passion fruit production. Rev. Biol. Trop. 60 (4): 1553-1565. Epub 2012 December 01.

The importance of plant diversity in maintaining the pollinator bee, Eulaema nigrita (Hymenoptera: Apidae) in sweet passion fruit fields Animal pollination is one of nature's crucial ecosystem services.The majority of wild plant species and even many crops depend on animals to be pollinated (Crane & Walker 1984, Daily 1997, Nabham & Buchmann 1997, Kearns et al. 1998, Kremen et al. 2007, Ricketts et al. 2008, Ollerton et al. 2011).According to Roubik (1995), approximately half of all pollinators of tropical plants are bees, which improve the pollination of at least 72% of 1 330 crop species.Among bees, those belonging to the tribes Centridini and Euglossini can be considered key pollinators in tropical ecosystems (Schlindwein 2000).
In recent studies carried out in Uberlândia, state of Minas Gerais, and São Francisco do Itabopoana, state of Rio de Janeiro, Brazil, Gaglianone et al. (2010) reported that Epicharis flava Friese, 1900 (Centridini) was the most frequent pollinator of P. alata in both areas and Eulaema species were also considered as potential pollinators.According to those authors, E. nigrita visited the flowers of P. alata exclusively for nectar collection.This indicates that pollen, another resource necessary for adult survival and brood production, must be obtained from other plants as it was already observed to Xylocopa species, important pollinators of other Passiflora species, like Passiflora edulis f. flavicarpa Deg.(Silva et al. 2010a).
Conservation of natural areas at the surroundings of crops and the pattern of floral resources used by bees vary according to food plants' availability at a spatiotemporal scale (Kremen et al. 2007, Ricketts et al. 2008, Silva 2009).The knowledge of such patterns is of fundamental importance since it provides support to determine the plants that may be used in the conservation and maintenance of bee populations at the surrounding areas of passion-fruit crops.There are two methods to identify the species of plants used by bees in their diet: by direct observation of the bees in the flowers (Antonini & Martins 2003, Andena et al. 2005) and through identification of pollen grains present in brood cells, feces, or in the bees' body (males and females) (Silva et al. 2010a).During the collection of floral resources, both male and female touch the anthers leaving the pollen grains adhered to their body.So, the pollen grains deposited on the males' body indicate the nectar sources, while those deposited on the females indicated both, nectar and pollen sources.Therefore, the pollen grains are natural markers of floral resources used by bees as food for adults and immatures, as observed by Silva et al. (2010a, b).For that reason the pollinic analysis is an important tool in the investigation of the plants which maintain the bees.This study aimed to identify the species of native plants that are visited and used by E. nigrita in their diet during the flowering of sweet passion fruit.This information will be very important since they will indicate supplementary pollennectar sources which must be used for the conservation of the populations of E. nigrita in neighbouring areas to the crops.

Study area:
This study was carried out at Campo Alegre farm, Uberlândia-Minas Gerais State, during the flowering period of P. alata in the study region (October 2006-May 2007).The farm comprises a total area of 114ha, of which 20% is protected as reserve, one hectare is occupied by commercial cultivation of P. alata and seven hectares are occupied by Passiflora edulis f. flavicarpa.
The local climate has two well-defined seasons: a cool/dry season extending from March-September, and a hot/wet season extending from October-February.During the study period, the mean monthly temperature ranged from 20.0-26 o C, and the precipitation ranged from 11-84mm (Fig. 1).Records on temperature and precipitation were obtained from the Laboratório de Climatologia e Recursos Hídricos, da Estação Climatológica da Universidade Federal de Uberlândia, MG (Fig. 1).

Availability of floral resources in the surroundings of the Passiflora alata culture:
The surroundings of P. alata plantations at Campo Alegre farm were predominantly pastures and scattered patches of Cerrado stricto sensu.In order to evaluate the availability of floral resources potentially used by E. nigrita, we studied the floristic composition and the flowering phenology in a Cerrado reserve at "Clube de Caça e Pesca Itororó de Uberlândia" (CCPIU), located at 5 000m away from the studied plantation of P. alata.This distance between CCPIU and the studied area is within the flight range of many euglossine bees, which may forage at up to 23km from their nest (Janzen 1971).Native flowering plants were sampled along a transect of 1 000m long and 10m wide on each side, thus covering a total of 2ha of Cerrado flora, the dominant vegetation in the region studied.The transect was sampled once a month during the flowering period of P. alata (October 2006-May 2007), recording all flowering species as well as the number of individuals per species.The voucher specimens were deposited in the Herbarium Uberlandense at the Federal University of Uberlândia, Brazil.
Collection of bees for sampling pollen grains: For the collection of pollen grain samples from bees, we captured specimens of E. nigrita once a month during the flowering period of P. alata from 8:00am-4:00pm (126 sampling hours in total).During that period, the bees were collected in the flowers of P. alata, for 15min per hour.This method was adopted in order to not interfere with pollination and fruit production in the study area, because it is a commercial area of P. alata.The bees were collected with an insect net and carefully verified in which part of the body (head, thorax, metasoma, corbiculae) the pollen grains were deposited.The pollen grains were sampled following the method suggested by Silva et al. (2010a), which consists of placing the bees into transparent vials containing 1mL of water in order to remove the pollen grains attached to the body of the individuals.The individuals were then released and 4mL of 70% alcohol were added in each vial.The material of each collected individual corresponded to one pollen sample, being analysed a total of 47 samples.

Pollen analysis:
From each plant species found in the surrounding area, we removed the anthers of at least four individuals in order to set up a reference collection of pollen grains, using acetolysis as proposed by Erdtman (1960).The same method was used for pollen grains samples of the bees.All prepared microscope slides were deposited at the Pollen Slide Collection of 'Laboratório de Morfologia vegetal, Microscopia e Imagem' (LAMOvI) at the Federal University of Uberlândia (UFU).From each bee pollen sample, we mounted three microscope slides that were qualitatively and quantitatively analyzed.For the qualitative analysis we compared the pollen grains removed from the individuals, with the pollen from plants sampled in the surrounding area (Pollen Slide Collection), and also with pollen grains from the region's Pollen List described by Silva et al. (2010b).For the quantitative analysis, we counted the first 400 pollen grains found on each microscope slide as suggested by Montero & Tormo (1990).From these, we determined the percentages following classification of Barth (1970) and Louveaux et al. (1970Louveaux et al. ( , 1978)): predominant pollen (>45% of total pollen grains on the slide), secondary pollen (from 15-45%), important minor pollen (3-15%) and minor pollen (<3%).In this study, those plant species whose pollen grains were found at percentages of more than 15% in the samples each month, were considered the preferably used plants for obtaining floral resources.In order to evaluate the importance of the native plant species in the surrounding area for the maintenance of E. nigrita populations, we did not consider the pollen grains of the two Passiflora species cultivated in the area in the quantitative analysis.
To determine the use of nectar by E. nigrita from flowers of species of native plants, floral traits were evaluated to check if the floral morphology favoured just the collection of nectar, for example, in species with tubular corollas which have nectariferous chamber or narrow floral tube.The flowers were also analyzed as determine the position of the anthers, to understand the process of visit, and if was possible, or not that E. nigrita also collect pollen beyond the nectar.Direct observations were done in the flowers on surrounding cultures whenever possible, to check the behaviour of E. nigrita in the flowers.Additional data were obtained in the literature.
Statistical tests were done using the statistical software BioEstat 5.0 (Ayres et al. 2005).Pearson correlation test was applied to assess the degree of association among climate (monthly mean temperature and rainfall) and availability of floral resources in the areas surrounding the P. alata plantation, and the number of bees sampled per month during flowering period of P. alata.

RESULTS
In the study area, flowering of P. alata occurred continuously from October 2006 to May 2007, although some peaks of open flowers were observed during the rainy season (October-February).At the beginning of the dry season (March-May) a decrease on flowering was observed.Thirty four females and 13 males of E. nigrita were collected.Of the sampled females, 24 individuals carried pollen on their corbiculae.Fifteen of these pollenloaded females were collected from 8:00am to 11:00am and nine of them from 12:00 noon to 4:00 pm.Female bees were most frequently observed on P. alata flowers in October, December, and March whereas males were most abundant in April and May (Fig. 2).No correlation was observed between the number of active bees on flowers and climatic conditions (r=0.43, p=0.28, for precipitation and r=0.13, p=0.75, for temperature).
During the collection of bees, we observed that pollen grains were distributed throughout the body of male individuals, but especially on the thorax (n=13); while in the body of females, the pollen grains were observed on the thorax, metasoma and predominantly in the corbiculae (n=34).Among the types of pollen grains accumulated on the thorax of males and females, that of P. alata was the predominant.In addition to the pollen of P. alata, we found pollen grains from 35 native plant species on females and from 14 species on males.There was an overlap between males and females in the visit of some plant species, being identified in total 40 plant species distributed in 32 genera and 19 botanical families, excluding the pollen grains of the passion fruit species (Table 1).
The use of floral resources by E. nigrita varied in the course of the studied period, with the highest number of pollen types in March, although in this month we observed one of the lowest numbers of flowering species in the plantation surroundings (Fig. 3).There was no correlation between the number of active individuals and the number of individuals flowering in the area surrounding the sweet passion-fruit plantation (r=0.40,p=0.33) (Fig. 3).However, the number of active bees was correlated with the number of pollen types identified in the bee samples (r=0.83,p=0.01) (Fig. 3).The amount of different pollen types carried by bees was strongly associated with the number of plant species providing pollen as pollinator-attractant (r=0.98,p<0.0001), but not with the number of species providing nectar (r=0.55,p=0.18).
The flowers visited by E. nigrita share several characteristics such as color, shape, type of anther, and size (Table 1).Those plant species represented 25% of the total of plants visited in each month.The most visited plant species to collect nectar have long corolla, which facilitates access to this flower resource by bees with long tongue, as E. nigrita.The species of plants that provide the pollen as attractive floral Frg: Floral resource gathered by E. nigrita.Sh: floral shape (tu: tube, di: dish, stan: standart, camp: campanulate); Sy: floral symmetry (act: actinomorphe, zig: zigomorphe); Co: color (li: lilac, re: red, ye: yellow, ni: no information, wh: white, pu: purple, or: orange, cr: cream); Ad: anther dehiscence (long: longitudinal, por: poricidal).
resources in this study have poricidal anthers, that are vibrated by females during visits for pollen grains removal.
In general, E. nigrita visited more native plant species to collect pollen (28 species) than to collect nectar (17 species) (Table 1).The pollen grains most frequently encountered in the bee pollen-samples were from Solanum lycocarpum St. Hil.(Solanaceae) and Rhynchanthera grandiflora (Aubl.)DC. (Melastomataceae).The flowers of these two plant species have poricidal anthers and provide only pollen as floral resource.From the set of species used by E. nigrita, S. lycocarpum and R. grandiflora were preferably visited during their flowering periods as indicated by the frequency of pollen grains present in the bee pollen samples (Table 1).Between the flowering season end of S. lycocarpum and the beginning of R. grandiflora (January and February), other species, such as Trembleya parviflora (D.Don) Cogn.(Melastomataceae), Myrcia guianensis (Aubl.)DC. (Myrtaceae) and Cambessedesia hilariana (Kunth) DC. (Melastomataceae), were the most visited.These three species were visited and/or used as source of floral resources only when S. lycocarpum and R. grandiflora were not bloom (Fig. 4).Concerning nectar collection, with the exception of P. alata that was predominant in all samples, we observed that Crotalaria micans Link (Fabaceae) in April (32.9%), and Vochysia tucanorum Mart.(vochysiaceae) in November (25.0%), were the plants more visited by E. nigrita (Table 1).
Eulaema nigrita females collect floral resources on a variety of plants that are not used by males.These males collect only nectar and floral fragrances, as males of other Euglossini species (Dodson et al. 1969, Dressler 1982, Ackerman 1985, Roubik & Hanson 2004).
The collection of high amounts of pollen and nectar by females of E. nigrita in the study site reflects, mainly, the demand for these resources to the provision of brood cells.Nectar plays an important role not only to feed the brood, but also as an energetic source to the maintenance of adults.The energy expenditures during female foraging for nest-building material and for the provision of cells are very significant due to female's performance of long trips to collect excrement, mud, resin and larval food (Pereira-Martins & Kerr 1991, Santos & Garófalo 1994).Euglossine bees collect nectar from tubular flowers and these flowers have nectar with a relatively low concentration of sugar (Roubik et al. 1995), leading females to extend the search for flowers in the field.On the other side, in an opportunistic way, E. nigrita may exhibit floral constancy in sources of nectar that are more concentrated and produced in high quantities, as it occurs in flowers of P. alata that exhibit on average a sucrose concentration of 39.65%±2.63 and a volume of 408µl±76.71per flower (Rocha 2006).According to Roubik et al. (1995), E. nigrita collects nectar at a concentration ranging from 20-49%.Another factor that may influence the collection of nectar in P. alata is the proximity of the nests to the cultivated area.As reported by Arriaga & Hernández (1998), the most important plants used by Euglossa atroveneta were near to the nests, and this may be related to energy saving in the flight.However, the pollen analysis indicated that the nectar produced by P. alata flowers is not enough to supply the necessary energy to maintain the population of E. nigrita.Thus, although P. alata has a relatively long flowering period, the samples revealed a wide variety of pollen types from nectarproducing native species, what corroborates the need to preserve the nectariferous plant species in the natural areas near to cultures, especially the species of Acanthaceae, Apocynaceae, Bignoniaceae, Bromeliaceae, Convolvulaceae and vochysiaceae.All plant species representatives of these families identified in this study, analysing the pollen grains removed from the body of E. nigrita, exhibit floral morphology that allows only the collecting of nectar by this bee.
It is important to emphasize again that P. alata flowers are used by E. nigrita exclusively to collect nectar and that these bees, as well as the others, need pollen for a balanced diet to survive (Weiner et al. 2010).From June to September E. nigrita depends on other sources than P. alata for the maintenance of nests, and even during its flowering period these bees still need other sources.Quantitative analyses of pollen grains in the samples were important because they revealed the floral constancy in species with poricidal anthers, such as the species of the families Fabaceae (subfamily Caesalpinoidae), Melastomataceae, Ochnaceae and Solanaceae.Pollen grains of these plants were included in the predominant category, corresponding to 12.5% of all species used by E. nigrita.This shows that these bees sporadically visit a wide variety of plants per month, but prefer a small fraction of them, for example S. lycocarpum and R. grandiflora.Roulston et al. (2000) observed that pollen grains from species with poricidal anthers exhibit high protein content (47.8%), and bees use pollen grains with a percentage between 12-61%.The protein directly influences the larval development of bees (Michener 2000, Minckley & Roulston 2006).This may justify the floral constancy of E. nigrita on S. lycocarpum and R. grandiflora, which blooming during several months throughout the year.The foraging movement between these two species may have occurred due to the variation in the percentage of flowering individuals.This reinforces the preference of E. nigrita females for plants with poricidal anthers.Similar data were observed for other bee species with vibration behavior, in which the predominant pollen grains were the ones from plant species with poricidal anthers, as for example Xylocopa spp.(Silva et al. 2010a) and Centris tarsata Smith, 1874 (Dórea et al. 2009).The pollen analysis enables to obtain quantitative data that are important because they confirm the fact that the plant species that are the most used by bees with vibration behaviour are the ones with more expensive resources, both in terms of protein content of pollen and in terms of energetic content of nectar.The quality of the floral resources must be taken into consideration when making management and conservation plans for pollinators, which has been declining with habitat alteration and reduction of ecological resources used for both feeding and nesting (Cane 2001, Ricketts et al. 2008, Potts et al. 2010, Bommarco et al. 2010).
Despite not being the most often pollinator of P. alata, E. nigrita is a potential pollinator for management in the plantations, since the main pollinator, Epicharis flava (Gaglianone et al. 2010), build their nest in soil, while E. nigrita nests in preexisting cavities.Regarding nesting sites, cement blocks with inside cavities used in constructions can be disposed in the field as trap-nests to attract nesting females of E. nigrita, as suggested by Garófalo et al. (2011).
Concerning feeding, the plants preferably visited and used by E. nigrita, which were identified in the present study, must be preserved or reintroduced in the proximities of P. alata to favour the maintenance of that species.The management of pollinators in the surroundings of P. alata is important, since this plant is a self-incompatible species (varassin & Silva 1999), and in the absence of pollinators, growers are forced to pay for hand pollination, which increases production costs (Pereira-vieira et al. 2010).Keeping pollinators in cultivated areas is still more feasible to ensure sweet passion fruit production.
species.Thanks to the anonymous reviewers for careful analysis and important contribution.

Fig. 1 .
Fig. 1.Climate data (monthly mean temperature and rainfall) for the flowering period of Passiflora alata, in Uberlândia, Southeastern Brazil.

Fig. 4 .
Fig. 4. Percentage of individuals of these most frequently plant species and constantly used by Eulaema nigrita in the surroundings of the plantation of Passiflora alata in Campo Alegre Farm, Uberlândia, Southeastern Brazil.The column left indicate the percentage of flowering individuals of the plant species more visited.The column right indicate the percentage of the pollen grains in this plant species samples.