Bees ( Hymenoptera : Apidae ) present in the flowers of the balsa wood Ochroma lagopus Swartz , 1788

The flower of balsa wood holds about 10 to 15 mL of nectar, which helps attracting pollinating agents, since the genus Ochroma is incapable of self-fertilization. However, a high mortality of bees is observed in these flowers. The present study investigated the frequency and constancy of mortality of the individuals of the family Apidae that fed on nectar from the balsa wood. Data was gathered from June to August 2008, in Lavras – Minas Gerais State, Brazil. In addition, the survival of the Africanized bees that fed on the nectar of this flower was compared to those that fed on 50% aqueous solution of honey. Forty flowers were analyzed, and 949 individuals of the orders Hymenoptera (98.1%), Hemiptera (0.95%), Coleoptera (0.74%) and Diptera (0.21%) were collected. Most Hymenoptera individuals were bees of the genera Partamona and Trigona (677 individuals), which were considered of constant occurrence. Flowers producing up to 16.7 nectar mL were found. The nectar diet contained 16.44% of total sugar, and resulted in low survival of the bees in laboratory (31.32  2.37 hours), compared to a diet of 50% aqueous solution of honey (112.32 2.03 hours).


Introduction
The flower of balsa wood, Ochroma lagopus Sw., is a plant belonging to family Malvaceae (SOUZA; LORENZI, 2005), which occurs naturally in the Amazon region, at altitudes varying from 0-1,000 m (LORENZI, 2002).Rainfall ranging of 1,500-3,000 mm, and temperature from 22 to 27°C are the requirements for its growing (CAMPOS; UCHIDA, 2002).This vegetal species grows best when there is intense sunlight, and can tolerate dry conditions very well, as long as air humidity levels do not go lower than 75% (MARENCO et al., 2001).More than 28 genera and 200 species of balsa wood grow in tropical and subtropical regions.The family Bombacaceae has great economic significance because it is used as a construction material for boats, buoys, among other uses (LORENZI, 2002).It is an extremely fastgrowing species that can reach up to 30 meters, with trunk diameters of 60-90 cm (PAULA et al., 1998).
The fruits of the balsa wood are formed by an elongated capsule that opens into five valves when the seeds are ripe, releasing cream colored tubes containing plenty of seeds.The flowers provide abundant food for bats, birds and several insects.Acta Scientiarum.Biological Sciences Maringá, v. 32, n. 4, p. 343-348, 2010 Flowering is seasonal and the flowers last one to two days, holding 10 to 15 mL of nectar per flower (PAULA et al., 1998).This amount of nectar helps attracting pollinating agents with high energy demands, because the Ochroma genus is unable of self-fertilization (GRIBEL et al., 1999).According to Opler (1983), the O. lagopus is pollinated by bats.In turn, Mora et al. (1999) suggest that a coati species is the pollinator of this Malvaceae species.
The most important pollinating agents belong to the class Insecta, mostly bees of the order Hymenoptera (SANTANA et al., 2002).
One third from the worldwide crop production comes from crops that depend on animal pollination, with the bees accounting for 38% of the pollination of flowering plants, but probably not including O. lagopus pollination (KERR et al., 2001).
As for the flower of balsa wood, this insect-plant interaction arouses debate.Paula et al. (1997) identified saccharidosis, glucose and fructose, besides 16 amino acids (lysine, histidine, aspartic acid, asparagine, threonine, serine, glutamic acid, glycine, alanine, cystine, valine, methionine, isoleukin, leukin, glutamine and phenylalanine) in the floral nectar of this plant.According to Haydak (1970), the amino acids that play an essential role in the nutrition of bees are: lysine, histidine, threonine, valine, methionine, isoleukin, leukin, phenylalanine arginine and tryptophan.Therefore, all the essential amino acids, except arginine and tryptophan, are contained in the nectar of balsa wood, so it is a very nutritious and attractive food for bees.
However, some researchers noticed a significant amount of bees found dead in the nectar of this plant in the State of São Paulo (NOGUEIRA- NETO, 2002), and in Viçosa, in the region known as Zona da Mata, in Minas Gerais State (PAULA et al., 1997).Consequently, new research was conducted to identify the substances that might cause nectar toxicity, such as phenolic compounds, alkaloids and non-protein amino acids.Paula et al. (1997) did not identify any toxic substance in flowers of balsa wood, or any morphological adaptations, such as resin or any sticky substance that might cause bees to be stuck in the plant and die.Adler (2000) reports that some nectar may be toxic to bees due to the presence of microorganisms that produce ethanol.According to Winston (1983), galactosis, manosis and rhamnosus can be toxic to bees or reduce their longevity.Meanwhile, analysis of nectar of balsa wood did not show the presence of the referred substances.
Severe mortality of bees that collected the nectar from balsa wood also occurred in plants at the campus of the Federal University of Lavras, in Southern Minas Gerais State.The present study aimed to investigate the frequency of dead insects found in the nectar of Ochroma lagopus flowers in the city of Lavras, Minas Gerais State, and the survival rate of Apis mellifera Linnaeus, 1758, in laboratory bioassays, when fed on the nectar of this flower, compared to bees fed on a honey-based diet.

Material and methods
The present study was conducted from June to August 2008, which is the flowering period of the balsa wood, and was based on the observation of some O. lagopus trees in the campus of the Federal University of Lavras (UFLA), in Minas Gerais, State, Brazil.The city of Lavras is located at 21°C 14' South latitude, 45°C 00' West longitude and at 910 meters altitude.According to Köppen climate classification, the predominant climate in the region is mild and temperate -mesothermal -rainy, with dry winter and an average temperature below 22°C.
One randomly chosen flower was collected every day, always at 5 p.m., when bees finish foraging, during 40 days.The flowers were taken to the Insect Biology Laboratory of the Department of Entomology, and the insects found inside the flowers were collected, counted and identified.Afterwards, the flowers were placed in an inverted position and the nectar was poured into a glass vial, free from contamination, and immediately quantified.
The bees were identified to genus, as recommended by Silveira et al. (2002), whereas the other insects were identified to order.
The constancy of individuals found dead in the flowers, that is, the percent of taxonomic categories found in the survey was determined (SILVEIRA NETO et al., 1995) by the equation: C = (P/N) x 100, where: C = Constancy, P = number of members of a given taxonomic category collected, N = number of individuals collected.The classes were established by the calculation of the confidence interval (CI) of the mean at a 5% significance level.The taxonomic category was considered to be 'constant' when the percent value of individuals of this category collected was higher than the upper limit of the confidence interval; 'accessory', when the percent value of individuals of the taxonomic category collected fell within the confidence interval; 'accidental' when the percent value of individuals collected was below the lower limit of the confidence interval (LUDWIG; REYNOLDS, 1988).
The presence of Varroa destructor Anderson and Trueman mite was assessed in A. mellifera bees collected and in nectar.Although this ectoparasite occurs worldwide, there is little information on its population fluctuation in Brazil (ANDERSON; TRUEMAN, 2000).
In order to investigate the causes of the survival of bees fed on nectar of balsa wood, and construct the survival function, newly emerged bees from Central Apiary were used and a colony was selected to provide individuals for laboratory trials.
In the Insect Biology laboratory -DEN/UFLA, groups of ten individuals were placed in a 15 cm wide x 10 cm high PVC cage, with a tulle net placed on top and an organza fabric on the bottom, and maintained in an acclimatized chamber at a 29  2°C, UR 70  10% and photoperiod of 12 hours.
The experiment was conducted in a completely randomized design with two treatments and ten replicates, and each experimental unit had two 2 mL eppendorf tubes, one for distilled water and the other for the diets, representing the treatments: 1) 50% aqueous solution of honey; 2) nectar of balsa wood.
In order to quantify the total sugar contents, reducers (monosaccharides) and non-reducers (disaccharides) in the nectar of balsa wood and the 50% aqueous solution of honey used in the laboratory trials, chemical analysis using the Somogyi-Nelson method was performed.This method is based on copper ion reduction where sugar is heated in alkaline solution of arsenic tartrate, producing a blue-colored compound quantified by spectrophotometry at 510 nm (SILVA et al., 2003).This analysis may help to justify the attractiveness of bees to nectar compared to 50% aqueous solution of honey, and was performed in the Analysis Laboratory of the Department of Food Science -UFLA.
The number of surviving bees was counted every 12 hours until 120 hours.Since the survival data is censored, Kaplan-Meier analysis was performed to determine the average life period for individuals and obtain a survival curve for each treatment (COLOSIMO; GIOLO, 2006).The Log-Rank test was performed to compare the treatments.

Results and discussion
Forty flowers were analyzed, and 949 dead individuals of the orders Hymenoptera (98.10%),Hemiptera (0.95%), Coleoptera (0.74%) and Diptera (0.21%) were found.Most Hymenoptera individuals were bees of the Partamona, Trigona and Apis genera (Table 1).No insect was found dead in the nectar of 20% of the flowers.Nevertheless, the average mortality was 23.7 insects per flower, and it is important to stress that 112 individuals were found dead in one flower, 63 of which belonging to the Partamona genus and 40 of the Trigona genus.It is possible that these Hymenoptera bees are more attracted to the nectar of balsa wood, which allows us to suggest that mortality is caused by some toxic element present in the nectar, or by asphyxia due to drowning in nectar (Figure 1), this latter cause is the most plausible due to the amount of nectar produced by O. lagopus flowers.Concerning the greater number of Trigona bees compared to the Apis genus, according to Morgado et al. (2002), in sunflower crops in Lavras (Minas Gerais State), Trigona spinipes were more abundant than A. mellifera in some periods of time, which may be an indication of the high number of individuals of this species in the region, or colonies in nests surrounding those areas.However the study from Gamito and Malerbo-Souza ( 2006) determined T. spinipe as an accessory species, and the A. mellifera as a constant species in orange tree flowers.
The Partamona, Trigona and Apis genera were found as constant, and the others were found to be of accessory incidence.Despite the constancy observed in these insects, they did not interfere with the pollination of this species.No category was found to be accidental, due to the high standard deviation observed.
One significant finding from this study is that the V. destructor mites was not found in any of the 219 A. mellifera bees collected, not even in the nectar collected, which may indicate a low scale fluctuation in the population of this parasite, during the period of the present study, and that this parasite in some stage of life cycle is a phoretic mites.
Concerning the life period, A. mellifera adult bees were found to have an average survival time of 31.32  2.37 hours when fed on nectar of balsa wood, which is significantly lower than the average survival time of 112.32  2.03 hours observed for individuals fed on 50% aqueous solution of honey (Figure 2).The functions of survival of A. mellifera bees fed on a diet of 50% aqueous solution of honey and on a diet of nectar of balsa wood were drawn (Figure 3).The treatments were found to be significantly different according to Log-rank test at significance level  = 5 %, with a p-value < 0.001.The lethal time TL 50 for 50% of the population fed on a diet of balsa wood nectar was estimated at only 25.03  0.32 hours, whereas for bees fed on a diet of 50% aqueous solution of honey it was longer than 120 hours.
As for the volume of nectar collected, some flowers were found to have none, while others had as much as 16.7 mL of nectar, and the analysis of the nectar collected showed a low sugar content (16.44%) compared to the honey collected in hives of A. mellifera from UFLA's Apiary (75.80%) or to Melipona honey (77.88%) (ALVES et al., 2005) (Table 2).Bioassays to assess the toxicity of O. lagopus nectar in Partamona and Trigona were conducted by Paula et al. (1997).However, no increase was found in the mortality rate of bees fed on nectar compared to the control bees fed on a saccharose solution.Another hypothesis to explain the high mortality of bees found in balsa wood nectar would be the drowning in the nectar, but this hypothesis is refuted by an experiment conducted by Paula et al. (1997) where nectar of some flowers was collected and replaced by almost identical concentrations of saccharose, and the authors observed that bees that fell on this solution managed to get out and fly again.On the other hand, when bees fell on balsa wood nectar they were unable to fly, and died within 30 seconds.However, it is worth stressing that the Honey Balsa wood Honey Balsa wood breathing tubes of bees are connected to the outside through spiracles in their exoskeleton and when these insects are foraging their breathing is increased, so their abdomen is rhythmically contracted to improve gas exchange (WINSTON, 1983).In this case, when bees dive deep into the nectar, which has a different viscosity from the solution tested by Paula et al. (1997), asphyxia may occur due to the presence of some types of amino acids that obstruct the spiracles, and lead these insects to die.The addition of amino acid to the solution proposed by Paula et al. (1997) was not yet tested.The most likely cause of death of bees is asphyxia by obstruction of air sacs and spiracles by one of the amino acids present in the balsa wood.

Conclusion
The dead insects found in the nectar of balsa wood belonged to the orders Hymenoptera, Coleoptera, Diptera and Hemiptera, with great predominance of Hymenopterous species.
The Partamona and Trigona genera were constant categories among the individuals collected in Ochroma lagopus flowers, in the city of Lavras, and were attracted by the great amount of nectar in these flowers.
The longevity of Apis mellifera adult bees fed on nectar of balsa wood flower was shorter than when these insects were fed on a 50% aqueous solution of honey.
The total sugar contents in the nectar of balsa wood was significantly lower than the one found in the 50% aqueous solution of honey, however the amount of reducer sugars was the same.

Figure 1 .
Figure 1.Flower of Ochroma lagopus Sw. with dead insects in the nectar.

Figure 2 .
Figure 2. Mortality of Apis mellifera bees along the time fed on a diet of 50% aqueous solution of honey and nectar of balsa wood.Temperature of 29 ± 2°C, UR 70 ± 10% and photoperiod of 12 hours.

Figure 3 .
Figure 3. Function of distribution of survival of Apis mellifera bees fed on the following diets: 50% aqueous solution of honey, and nectar of balsa wood.Temperature of 29 ± 2°C, UR 70 ± 10% and photoperiod of 12 hours.

Table 1 .
Number of insects found dead in flowers of Ochroma lagopus, from June to August 2008.

Table 2 .
Analysis of total sugars, reducers and non-reducers by the Somogyi-Nelson method.