Abstract
Pollen samples were collected in three different periods from 11 Melipona capixaba Moure & Camargo hives and analyzed with melissopalynological methodology. A total of 33 pollen types were identified, of which 23 genera and 15 families were identified. The following families showed the highest pollen richness: Fabaceae (7), Myrtaceae (3), Solanaceae (3), Arecaceae (2), Asteraceae (2), Euphorbiaceae (2), Melastomataceae/Combretaceae (2), Rubiaceae (2), and Sapindaceae (2). The most frequent pollen types (>45%) were Eucalyptus, which generated great similarities between the samples, except one in which the Tibouchina was predominant. Although the majority of the pollen types showed low percentage values, the results demonstrated that M. capixaba has taken advantage of the polliniferous sources available in the Atlantic Rainforest as well as in the “Capoeira” (brushwood, secondary forest) and “ruderal” (field) plants, probably implying its importance as a pollinator of the native flora and of the exotic species.
1. Introduction
Melipona Illger constitutes the genus of Meliponini tribe with the biggest amount of species. It occurs in the whole neotropical region, which is the most diversified in the Amazon basin [1]. M. capixaba (popularly known as uruçu-preta or uruçu-capixaba) is endemic to the Atlantic Rainforest where it is restricted to the mountainous area of Espírito Santo State in the municipalities of Domingos Martins, Conceição do Castelo, Venda Nova do Imigrante, Alfredo Chaves, and Afonso Cláudio embracing Pedra Azul State Park, which is protected by the government. M. capixaba was described by Moure and Camargo [2] who referred it to the Meliponini group of the Amazon region and designated by Rocha and Pompolo [3] to the same karyotype group of M. scutellaris Latreille. Experiments proved that the lack of anatomical or behavioural isolation mechanisms allowed the crossing of these species when they were brought into the same area. These observations suggest that the two species are capable of forming fertile hybrids [4]. The fact that two ecologically different species of stingless bees, separated by more than 300 km, could still cross when placed in the same area suggests that there has not been any pressure to develop reproductive isolation [4]. However, M. capixaba is mentioned on the list of Endangered Brazilian species (Normative Instructions no. 3, May 27th 2003, Ministry of Environment) because its original habitat has been almost completely fragmented to make room to pastureland, coffee (Coffea sp), and Eucalytus sp cultivars. Currently only 10% of the original Atlantic Rainforest remains and has become a “hot spot” for this bee [4, 5].
Despite its endangered status and ecological importance as pollinator, few studies have examined the ecology and biology of M. capixaba. Pollen analysis may be useful to indirectly determine its food sources and help clarify its role as a vegetation pollinator [6]. Knowledge of the several pollen sources that are used by M. capixaba in its natural environment helps the beekeepers to manage them. Likewise, complementation of the ecological data, obtained through melissopalynological analysis, is important for the development of the preservation programs for this bee. This may help in directing the efforts to recover the vegetation in the affected areas utilizing botanic species that guarantee its food supply. There is only one publication in the scientific literature that reports a case of the workers of M. capixaba carrying pollinarium attached to the scutellum, of the orchid subtribe Maxillariinae species, possibly of the genus Maxillaria sensu lato or Xylobium [7]. However, food storage in the beewax pots of M. capixaba has not yet been palynologically analyzed. The aim of this study was to investigate the influence of the local flora on the pollen harvest by M. capixaba to characterize the vegetation in which the sources were obtained in order to help in the conservation efforts of this bee.
2. Materials and Methods
Pollen samples were collected from 11 hives of Melipona capixaba found inside the tree-trunks and maintained by beekeepers of three different regions. The pollen samples were collected in different periods (October 2007, May 2008, and March 2009) directly from their food storage in the beewax pollen pots. Six samples originated from Domingos Martins municipality were defined as JV1, JV2, JV3, EM4, JV9, and JV10; three from Alfredo Chaves (JOV5, JOV6, and JOV7), and two from Venda Nova do Imigrante (FC0, FC8), all of them from Espírito Santo State, Brazil (Figure 1, Table 1). The meliponary located in Venda Nova do Imigrante municipality (FC) stays near to the urban area, being a small farm where they raise chicken and cultivate orchids, besides raising stingless bees. The bees collect floral resources in small forest fragments nearby, and there are some Eucalyptus cultivars a few meters far from the meliponary. The meliponary from Alfredo Chaves municipality (JOV) is located within a Particular Reserve of the Natural Patrimony (a private land), an area with secondary forest that has been acquired by the land owner for ecological tourism. There are a lot of coffee and Eucalyptus cultivars in the region. The meliponary located north Domingos Martins municipality (JV) is inside a small farm near small forest fragments that are permanent forest reserves inside private properties in the neighborhood. There are fruit, vegetable, and Eucalyptus cultivars. The meliponary south Domingos Martins municipality (EM) is inside a small farm near great areas of primary or secondary native forests in high stages of succession, within a permanent forest reserve in “Hotel Monte Verde”, occupying an area of about 3,000 hectares, and that uses the place for ecological tourism. Besides that, this meliponary is close to two State Parks, State Park of Forno Grande and State Park of Pedra Azul. There are many family farms in the region that cultivate flowers, fruits, and vegetables, besides agro-ecotourism, without Eucalyptus cultivars.
All the sediments of one pollen pot were analyzed as a single samples of one hive. The samples were acetolysed [8]. Microscope slides were prepared using jelly glycerine and sealed with paraffin. All samples were observed under traditional light microscopy. Pollen identification was done using literature data [9, 10] and the reference pollen slide collection of the Palynology Research Center, Instituto de Botânica, from the Environment Department of São Paulo State. A total of 500 pollen grains of each sample were counted for the frequency calculations, as shown in Table 2. Percentage values above 2% of the total count are specified. Interpretation of the data takes in account all the pollen grains of plant taxa. The term “pollen type” means a single plant species or a group of species, or higher taxa, presenting similar pollen morphology. “Monofloral” means originating mostly from a unique plant species (pollen type with >90% of the total) [11].
The multivariate analysis was performed through Principal Component Analysis (PCA) in order to verify the pollen grain occurrence in the samples. The matrix comprised the absolute value of all taxa found in each sample. The absolute numerical variables were transformed into natural logarithms using the FITOPAC program [12] and the ordination was done through covariance matrix using PC-ORD 4.0 [13].
3. Results
Thirty-three pollen types were identified with regards to family, genus, and species taxa (Table 2). The families that showed the highest richness of pollen types were Fabaceae (7), Myrtaceae (3), Solanaceae (3), Arecaceae (2), Asteraceae (2), Euphorbiaceae (2), Melastomataceae/Combretaceae (2), Rubiaceae (2), and Sapindaceae (2). The variability between the samples of pollen comprised 82.7% on the two first axis in the Principal Component Analysis (PCA) (Figure 2). Axis 1 alone comprised 71.0%. Regarding correlation among the pollen types, the pollen samples showed great similarity related to the high occurrence of the Eucalyptus (Figure 3(e)), except for the EM4 sample, from Domingos Martins, in which it was absent. Further Tibouchina (Figure 3(d)) was common in the samples, absent only in the JV3 sample, but predominant in the EM4 sample (Table 2). Myrcia (Figure 3(f)), even though is also a main characteristic component, occurred in small percentages in all samples. Paullinia (Figure 3(h)) was observed in most samples, however in small quantities. Senna (Figure 3(b)) occurred in 5 samples with varying percentages. Other pollen types that were expressive, although present in few samples, were Alchornea (Figure 3(a)), Combretum (Figure 3(c)), Euterpe/Syagrus, Faramea, Piper (Figure 3(g)), and Solanaceae type 2 (Figure 3(i)) (Table 2).
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
(i)
4. Discussion
The species found in the State Park of Pedra Azul forest [14] include Alchornea triplinervia, Andira sp, Annona sp, Astronium graveolens, Cariniana estrellensis, Carpotroche brasiliensis, Cedrela sp, Didymopanax morototoni, Erythoxyllum subssesilis, Euterpe edulis, Faramea sp, Fuchsia regia, Geonoma schottiana, Guapira opposita, Melanoxylon sp, Miconia inaequidens, M. latecrenata, Myrsine coriaceae, M. parvifolia, M. umbellata, Nectandra sp, Ocotea sp, Rollinia sp, Senna sp, Serjania sp, Schizolobium sp, Solanum sp, Solanum capsicoides, Sorocea ilicifolia, Tabebuia sp, Tibouchina sp, and several species of the Myrtaceae family (mainly Myrcia and Eugenia genera), among others. In the forest outskirts one can frequently observe Guapira opposita, M. latecrenata, Erythroxylum ovalifolium, Psychotria hancornifolia, and M. coriacea. The areas of rocky outcrops present a singular herbaceous-shrubby stratum with Aechmea sp, Alcantarea imperialis, Baccharis sp, Epidendrum denticulatum, Fuchsia regia, Leandra sp, Melinis minutiflora, M. inaequidens, M. latecrenata, Poliavana sp, Polypodium sp, Pseudolaelia vellozicola, Tibouchina sp, Vellozia sp, Vernonia sp, and Vriesea carinata. The glades are characterized by Cecropia sp, Mimosa sp, Solanum sp, and Solanum capsicoides, with significant populations of the pteridophyta Pteridium aquilinum and Melinis minutiflora grass. In the urban gardens are found bromeliads and orchards (with guava, lime, orange, etc.). Pasture lands are dominated by Baccharis dracunculifolia, Bidens pilosa, Borreria verticilata, Chamaessyce prostrata, Chamaecrista sp, Crotalaria claussemi, Emilia sonchifolia, Mellines minutiflora, Panicum maximum, and Sida sp, among others. Based on this, it is confirmed that M. capixaba harvested the polliniferous resources from natives trees, shrubs, and vines (Alchornea, Arecaceae, Cassia, Combretum, Eugenia, Euterpe/Syagrus, Faramea, Myrcia, Myrsine, Paullinia, Senna, Serjania, several Solanaceae, and Tibouchina), as well as pollen types from native trees not present in the referred list (Anadenanthera, Aparisthmium, and Cydista heterophylla). The altered vegetation “Capoeira” (brushwood, secondary forest) and ruderal (field) plants were represented by the pollen types Cecropia, Combretum, Crotalaria, Desmodium, Eupatorium, Mimosa scabrella, Piper, Sida, several Solanaceae, Tibouchina and Vernonia, apart from the exotic species of Eucalyptus, and Rudgea jasminoides. In spite of the high richness of pollen types, the results demonstrated a similarity between the hives regarding the preferences of pollen of Eucalyptus, a widely cultivated tree in the region, and, with less intensity, pollen of Tibouchina, a common plant in the native forest. The EM4 sample from the meliponary south Domingos Martins was the only one that did not show Eucalyptus pollen grains, because it is located in one of the most well-preserved areas of the region, without cultivars of this plant in the surroundings. The other areas of the meliponaries (JOV, JV, and FC) are more deforested and replaced by pastures, and are near Eucalyptus cultivars. The palynological results from this sample showed that in the presence of native floral resources M. capixaba efficiently visits and harvest those resources, indicating their original pollen sources.
Studies based on the visitation of the pollinators of the flowers of the Atlantic Rainforest concluded that stingless bees are not specialized visitors in certain plant species. Only 7% of the plants in the Atlantic Rainforest are intensely visited by native bees; 77% of the plants are visited with less frequency, and 16% are not visited at all [15]. Analysis of palynological results in Brazil point to a great variety of trophic resources and generalized habits of pollen harvesting by the Meliponini that commonly visit a larger number of plant species [16]. However, regional studies show that the Meliponini frequently concentrate the pollen harvest to a few floral sources [17, 18]. Several studies under natural conditions in the Tropical Atlantic Domain concluded that colonies of different species of Melipona frequently searched floral sources of the trees of families Melastomataceae, Mimosaceae, Myrtaceae, and Solanaceae, apart from the genus Cassia of family Caesalpiniaceae [17, 19–21]. The present study corroborates in great part this tendency as the families Melastomataceae and Myrtaceae were the most procured families by M. capixaba for pollen harvest while families Mimosaceae, Solanaceae, and plants of the genus Cassia (or Senna), although utilized, were visited to a significantly lesser extent.
Selectivity or floral preferences is a behaviour that is frequently observed among M. scutellaris [22]. Field research in the northeast Brazil confirmed that it prefers trees of the Atlantic Rainforest, as well as “Capoeira” vegetation, to herbs, and that it is rather selective with reference to the food source [23]. Palynological studies on pollen pellets of M. scutellaris in forest areas showed that the floral source preferences depended on the plant species, demonstrating harvesting selectivity as follows, with decreasing order of importance: Myrtaceae, Mimosaceae, Anacardiaceae, Sapindaceae, Caesalpiniaceae, and Fabaceae [6]. In our analyses, the selectivity also occurred in relation to the M. capixaba with the Myrtaceae and the Melastomataceae families as the most popular pollen suppliers. Four pollen samples were monofloral and seven samples were heterofloral with a major percentage contribution of pollen types from these families. During the analyzed periods (March, May, and October) there was a preference of Eucalyptus and Tibouchina, in spite of other pollen sources in the region. Combretum (in one sample), Myrcia (two samles), Senna (two samples), and Solanaceae (one sample), also contributed with high pollen percentages. Other plants, including herbs, were sporadically visited by M. capixaba.
The search for M. capixaba in Eucalyptus is not a new fact for the Meliponini as the occurrence of its nectar and its pollen has been reported dominant or important in the pollen spectrum of honey and pollen pellets of other species of Melipona [18, 19, 21, 24, 25]. However, the almost exclusive usage of only one floral resource by M. capixaba can put this species in danger on the absence of this resource, besides the fact that it cannot offer all the nutritional necessities needed by the bees. Other research demonstrated that the pollen nutritional composition harvest by Apis mellifera did not show a correlation with the pollen type diversity, but showed a correlation with the predominance of specific pollen types [26, 27]. These authors pointed as well that the different floristic compositions have influenced on pollen pellet quality, and the harvest in different food sources by the bees is important for obtaining a well-balanced diet. The fact that the predominant polliniferous source for M. capixaba was Eucalyptus demonstrates as well an important economic issue. Nowadays there is some interest in local Eucalyptus cultivars, but this is not constant, because it depends on market interest. If in the future Eucalyptus cultivars would be replaced by any other reason (environmental or economic importance) and enough native pollen sources to feed the colonies would be absent it would have a significant prejudice for the bees, because M. capixaba is restricted to this montane region within the Espírito Santo’s Atlantic Forest. Its restriction to this small occurrence area can be evolutionary related to the local biological characteristics, as local native flora. The palynological analysis showed the main pollen sources of native plants for M. capixaba and thus provided important information for future researches on the pollination biology, allowing decision making on the creation of sustainable pastures to these bee.
In conclusion, M. capixaba took advantage of the polliniferous sources in the forest, especially Combretum, Euterpe/Syagrus, Faramea, Myrcia, Senna, and several species of Solanaceae and Tibouchina, indicating its importance as a pollinator of the native flora, as well as in ruderal plants (“Capoeira”), even though the characteristic pollen types of these environments occurred in low percentages. The main pollen harvest was in Eucalyptus cultivars. The mechanisms that increased the frequency of utilization of Eucalyptus, even in natural forest areas, still have to be better understood in order to clarify the influence of the nutritional quality of this pollen on its diet. Physiochemical analysis of the pollen from the Myrtaceae family showed a positive and highly significant relation of crude proteins in Eucalyptus and, consequently, a complementary negative relation for total carbohydrates [26, 27]. Studies should be undertaken to verify the influence that a diet almost exclusively consisting of Eucalyptus has on M. capixaba.
In order to guarantee a varied supply of M. capixaba pollen, it is recommended that the beekeepers and others interested in its preservation recover affected areas, reforesting with polliniferous native plants that now are known. The foraging behaviour of the M. capixaba explains important ecological consequences in terms of persistence capacity of the species in the environment, as in cases of local extinction of their preferential native floral sources.
Acknowledgments
The authors acknowledge the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) for the fellowship of “Produtividade em Pesquisa” (Process number 301220/2009-3) to the first author and to FAPEMIG for financial support. They are grateful to the beekeepers Ernesto Mazzoco, Fábio Calimã, José Valim, and Jovane R. Carvalho for providing the samples. Thanks are due to Erik Bernhard Lidgren for the translation.