Differential insecticide susceptibility of the Neotropical stingless bee Melipona quadrifasciata and the honey bee Apis mellifera
- First Online:
- Cite this article as:
- Del Sarto, M.C.L., Oliveira, E.E., Guedes, R.N.C. et al. Apidologie (2014) 45: 626. doi:10.1007/s13592-014-0281-6
- 460 Views
The toxicity of three insecticides frequently used in Neotropical tomato cultivation (abamectin, deltamethrin, and methamidophos) was estimated on foragers of the Neotropical stingless bee Melipona quadrifasciata (Lep.) and the honey bee Apis mellifera (L.). Our results showed that the susceptibility varied significantly with the type of exposure (ingestion, topical, or contact), and there were significant differences between species. While M. quadrifasciata was usually more susceptible to insecticides (except for abamectin) in realistic exposures (via ingestion and contact) than A. mellifera, the former was less susceptible than A. mellifera to topically applied insecticides, a less realistic means of insecticide exposure. These findings challenge the common extrapolation of toxicity bioassays with A. mellifera to all (native) bee pollinators. Such equivocated extrapolation may compromise the significant services provided by native bees in Neotropical ecosystems.
Keywordsinsecticide exposureacute toxicitybuzz pollinatorswild bees
The reported decline of bee populations and the potential impairment of the sustainability of pollination services performed by these insects are the target of current worldwide concerns (Potts et al. 2010; Cameron et al. 2011; Lautenbach et al. 2012). Habitat destruction, climate change, pathogens, and pesticides are thought to be the main contributors to colony decline of the honey bee Apis mellifera (L.) (Biesmeijer et al. 2006; Potts et al. 2010; vanEngelsdorp and Meixner 2010), and recent findings suggest that multifactor interactions between honey bee diet, parasites, diseases, and pesticides potentiate the decline in managed honey bee colonies (vanEngelsdorp and Meixner 2010; Cornman et al. 2012; Becher et al. 2013; Pettis et al. 2013).
The significant decline of managed honey bee colonies that has been reported mainly in the USA and Europe drew further attention to wild pollinator communities and their potential importance for pollination services in certain landscapes and crop systems (Winfree et al. 2007; Garibaldi et al. 2013; Jha and Kremen 2013). Native pollinator bees can perform equal to or better than the honey bee as pollinators for some crops (Maeta and Kitamura 1981; Freitas and Paxton 1998), significantly contributing to crop production even when honey bees are present (Winfree et al. 2007; Garibaldi et al. 2013). The stingless bees of the genus Melipona have been recognized as important pollinators of native plants in subtropical and tropical areas (Antonini et al. 2006) and have been recently recognized as promising agents for commercial pollination in crop systems, such as those of tomatoes (Del Sarto et al. 2005; Bispo dos Santos et al. 2009), eggplants (Nunes-Silva et al. 2013), and sweet peppers (Cruz et al. 2005). However, considering that pesticide application is a common agricultural practice in tropical areas (e.g., 500 pesticide active ingredients are registered in Brazil (Silveira and Antoniosi-Filho 2013)), and that over 150 pesticides are currently used and known to be toxic to honey bees (Devillers et al. 2003; Johnson et al. 2010; Mullin et al. 2010), it is likely that stingless bee species that provide pollination services may also be the non-intended target of harmful insecticide effects.
Despite the recognized ecological and agricultural importance of stingless bees in tropical areas, the majority of studies assessing insecticide impacts on pollinators have focused on honey bees with few studies assessing the insecticide susceptibility of stingless bee species (Moraes et al. 2000; Valdovinos-Núñez et al. 2009; Lourenço et al. 2012; Tomé et al. 2012). Here, we compared the susceptibility between foragers of the honey bee A. mellifera and of the Neotropical stingless bee Melipona quadrifasciata (Lepetelier) to three insecticides (deltamethrin, methamidophos, and abamectin) that are commonly used against insect pest species in Brazilian tomato fields (MAPA 2013). Concern regarding the stingless bee M. quadrifasciata is due to its wide distribution in Brazil, especially where field and protected tomato systems are cultivated (Del Sarto et al. 2005; Bispo dos Santos et al. 2009). Furthermore, M. quadrifasciata belongs to the same genus as the bee species Melipona capixaba (Moure and Camargo), which is recognized as an endangered species not only by the Brazilian Ministry of Environment (Normative Instructions no. 3, May 27, 2003 (Resende et al. 2008; Luz et al. 2011)) but also by the International Union for the Conservation of Nature and Natural Resources (IUCN 2013), reinforcing the importance of assessing insecticide impacts on M. quadrifasciata. Insecticide bioassays were also performed with the honey bee because of its broadly recognized role as the bee pollinator model in ecotoxicology protocols for assessing the toxic effect of pesticides on pollinators (Felton et al. 1986).
Insecticide toxicity assessments using different means of exposing the stingless bee M. quadrifasciata (in addition to honey bees) provide basic toxicological information to guide insecticide use and minimize their potential non-target impact on native pollinators. The findings of this study also assist in assessing the validity of the honey bee as an indicator of insecticidal impact on native stingless bees.
2 Material and methods
All insects used in this investigation were obtained from seven colonies of the Neotropical stingless bee M. quadrifasciata or from seven colonies of the Africanized honey bee A. mellifera maintained under field conditions at the Experimental Apiary of the Federal University of Viçosa (UFV, Viçosa, MG, Brazil, 20°45′ S, 42°52′ W). To ensure genetic variability between the colonies and to obtain more reliable toxicological estimates, sets of foragers from different colonies of either bee species were considered as replicates.
The concentration–mortality bioassays followed directive numbers 213 and 214 of the Organization for Economic Cooperation and Development (OECD 1998a, b). The insecticides used are neurotoxic compounds registered and commonly used for pest control in field and in protected tomato crop systems in Brazil (MAPA 2013). Attempting to mimic a more realistic insecticide exposure in the field, bees were orally exposed to the insecticide commercial formulations (pyrethroid deltamethrin: Decis 25 EC, Bayer CropScience, São Paulo, Brazil; organophosphate methamidophos: Tamaron BR, Bayer CropScience, São Paulo, Brazil; avermectin abamectin: Vertimec 18 EC, Syngenta Proteção de Cultivos LTDA, São Paulo, Brazil). For the assays in which the bees were topically exposed or contact-exposed, technical grade insecticides (purity ≥ 90 %) were used and directly obtained from the manufacturers (deltamethrin and methamidophos: Bayer CropScience, São Paulo, Brazil; abamectin: Syngenta Proteção de Cultivos LTDA, São Paulo, Brazil). Whenever suitable, the insecticides were classified according to their toxicity to bees as described by Felton et al. (1986) in the following categories: highly toxic (median lethal dose (LD50) values <1.0 μg a.i./bee), moderately toxic (1.1 < LD50 < 10 μg a.i./bee), slightly toxic (10.1 < LD50 ≤ 100 μg a.i./bee), and virtually non-toxic (LD50 values >100 μg a.i./bee).
2.3 Insecticide susceptibility bioassays
The insecticide susceptibility of bee foragers was evaluated using three means of exposure: ingestion (in honey syrup 50 % v/v), topical, and contact with dry residues. In all cases, preliminary concentration–mortality bioassays were performed to determine the concentration range to be used in the bioassays (i.e., the interval between the highest concentration for which no mortality was observed and the smallest concentration for which 100 % of the tested bees were killed). Once the concentration range was recognized, six to eight concentrations were established for the definitive concentration–mortality bioassays. In the control treatments, the bees received only honey syrup (oral exposure), were topically exposed to the solvent (acetone), or were exposed to a solvent-treated surface (free of insecticide residue).
2.3.1 Ingestion bioassays
For these bioassays, M. quadrifasciata bees were collected using a 10-cm-long plastic tube (diameter of 20 mm) connecting the colony entrance to a wooden cage (9 × 9 × 3 cm). The bees were arrested on the wooden cage as they left the colony. To collect the workers of A. mellifera, these wooden cages were placed at the hive entrance. In each wooden cage, a 1-cm-diameter hole was drilled in one of the cage walls, and a 2-mL Eppendorf tube was inserted. At the bottom of this tube, a small hole (~0.5 mm) was drilled to allow the bees (group of seven) access to the honey syrup diet. Insecticide solutions were mixed in the diet and never exceeded 10 % of the total diet volume. The bees were fasted for 1 h prior to the experiments before allowed access to the insecticide-treated diet for the subsequent 5 h. The food consumption during this 5-h period was calculated by subtracting the amount of food left in the Eppendorf tube, which allowed us to estimate the dose of insecticide to which the insects were exposed. After this 5-h period, the bees were allowed access to an insecticide-free diet, and mortality was recorded after 24 h. The bees were considered dead if unable to walk when prodded with a fine hair brush. Seven replicates (i.e., a wood cage containing seven bees from different colonies of the same species) were used for each insecticide concentration. During the experimental period, the cages were maintained at 28 ± 2 °C and 65 ± 5 % relative humidity (r.h.).
2.3.2 Topical bioassays
In these bioassays, the bees were ice-chilled prior to the application of the technical grade insecticides, which was performed by applying a 1-μL insecticide solution to each adult forager bee with an automatic microapplicator (Burkard, Rickmansworth, UK). Acetone was the solvent used in the application of the technical grade insecticides. The insecticide was applied on the ventral side of the region located between the second and third pair of legs. The bees were subsequently placed in a 9-cm-diameter Petri dish with the bottom covered with filter paper. Each Petri dish received seven foragers of the same bee species. The bees were kept on the Petri dishes for 24 h with a honey syrup diet offered ad libitum. Seven replicates were used for each insecticide concentration. Bee mortality was recorded 24 h after topical insecticide application. The bees were considered dead if unable to walk when prodded with a fine hair brush. During the experimental period, the Petri dishes were maintained at 28 ± 2 °C and 65 ± 5 % relative humidity (r.h.).
2.3.3 Contact bioassays
Technical grade insecticide solutions (1 mL; acetone was used as a solvent) were applied to the filter papers (Whatman no. 1), which were left to dry for 30 min. These filter papers were used to cover the bottom of Petri dishes (9.0-cm diameter) that had their inner walls coated with Teflon® PTFE (DuPont, Wilmington, DE, USA) to prevent the bees from escaping. Each Petri dish received a group of 10 adult forager bees from different colonies of the same species. Insect mortality was recorded after 24 h of exposure, and the bees were considered dead if unable to walk when prodded with a fine hair brush. An insecticide-free honey syrup diet was provided ad libitum to the bees during the exposure. The bioassays were performed at 28 ± 2 °C and 65 ± 5 % relative humidity (r.h.).
2.4 Statistical analysis
The concentration–mortality data were subjected to probit analysis (PROC PROBIT; SAS Institute 2008). The differential insecticide susceptibility between M. quadrifasciata and A. mellifera was calculated for each insecticide based on the estimated LD50 (or median lethal concentration (LC50)) for each insecticide and bee species and dividing the LD50 (or LC50) value obtained for M. quadrifasciata by the LD50 (or LC50) value obtained for A. mellifera (Robertson and Preisler 1992). The 95 % confidence limits of these toxicity rate estimates were considered to be significantly different (P < 0.05) if they did not include the value 1 (Robertson and Preisler 1992).
Insecticide susceptibility was assessed in forager bees of M. quadrifasciata and A. mellifera (the model pollinator species). The insecticide susceptibility varied significantly with the type of exposure (ingestion, topical, or contact), and there were significant differences between species for some means of insecticide exposure.
3.1 Susceptibility to ingested insecticides
Relative toxicity of ingested insecticides (commercial formulations) to M. quadrifasciata and A. mellifera.
Number of samples
Slope ± S.E.
LD50 (95 % FL)
LD95 (95 % FL)
TRb (95 % CLc)
4.7 ± 0.50
6.1 ± 0.99
2,7 ± 0.35
1.8 ± 0.44
3.7 ± 0.47
9.6 ± 2.37
3.2 Susceptibility to topically applied insecticides
Relative toxicity of topically applied insecticides (technical grade) to M. quadrifasciata and A. mellifera.
Number of samples
Slope ± S.E.
LD50 (95 % FL)
LD95 (95 % FL)
TRb (95 % CLc)
3.0 ± 0.39
6.7 ± 1.54
3.0 ± 0.38
3.3 ± 0.47
2.0 ± 0.44
2.5 ± 0.48
3.3 Contact susceptibility to dry insecticide residues
Relative toxicity of contact-exposed insecticides (technical grade) to M. quadrifasciata and A. mellifera.
Number of samples
Slope ± S.E.
LC50 (95 % FL)
LC95 (95 % FL)
TRb (95 % CLc)
2.8 ± 0.42
2.3 ± 0.34
2.3 ± 0.25
4.1 ± 0.81
3.9 ± 0.73
3.0 ± 0.47
The present study assessed the insecticide susceptibility of the Neotropical stingless bees M. quadrifasciata and the honey bee A. mellifera. The insecticides exhibited low overall toxicity when topically applied to adult forager bees, while the ingested insecticides were usually very harmful to both bee species. Furthermore, contact exposure to dry residues of deltamethrin and abamectin was moderately harmful, while methamidophos was virtually non-toxic to both bee species. Our results revealed that M. quadrifasciata is usually more susceptible to insecticides in realistic exposures via ingestion and contact than A. mellifera, with the exception of abamectin. In contrast, M. quadrifasciata is usually less susceptible to topically applied insecticides, which is a less realistic means of insecticide exposure. The findings reinforce the importance of testing ingestion and contact exposure rather than only topical exposure and challenge the notion that the honey bee is a suitable pollinator model (or bioindicator) for insecticide toxicity and risk assessment.
Insecticide application is very common in agroecosystems, which may harm many insect species that assist in pollination (Freitas et al. 2009; Johnson et al. 2010; Krupke et al. 2012). Studies with pyrethroids, fipronil, and especially neonicotinoids have provided accumulated evidence that insecticides can cause potential problems for colonies of A. mellifera (Cresswell et al. 2012; Gill et al. 2012; Tirado et al. 2013). However, the risk posed by insecticides on other pollinator species has been largely unexplored, although suspicions that these native pollinators are seriously threatened by insecticide applications have begun to arise (Morandin et al. 2005; Thompson et al. 2007; Osborne 2012; Biddinger et al. 2013; Decourtye et al. 2013).
Bee exposure to insecticide may take place through a variety of means. Although pesticide exposure may occur with compounds applied to control Varroa parasitic mites in beehives, pesticide exposure usually results from the contact between bee foragers and contaminated plant surfaces, the harvesting of contaminated pollen and nectar, or the ingestion of contaminated sap from plants originating from insecticide-coated seeds (Johnson et al. 2010; Gill et al. 2012; Mullin et al. 2010). In our study, we assessed insecticide toxicity in pollinator bees and explored distinct means of exposure (ingestion, topical, and contact exposure) by using highly controlled conditions and rigorous experimental design bioassays that enabled comparative toxicity assessments between Neotropical stingless bees and A. mellifera. The results obtained here revealed that ingested insecticides are very harmful for both bee species but are 10-fold more effective in killing native species than the honey bee. This is in accordance with previous studies that have shown that other native bees are more susceptible to insecticides (Cresswell et al. 2012; Lourenço et al. 2012; Hardstone and Scott 2010). However, the insecticide abamectin provided contrasting results, showing that the native stingless bee is more tolerant than the honey bee for ingested and topically applied abamectin.
Different insecticide susceptibility may result from insecticide- or insect-related differences. For instance, insecticide formulations usually enhance the insecticidal activity of the active ingredient (i.e., the insecticide), and the Africanized honey bees used in our study, which prevails in Brazil, may respond differently to insecticide exposure than the European honey bee, which prevails in most of the USA and Europe. Indeed, the DL50 values for topical insecticide applications on A. mellifera from our study are higher than those reported by the US Environmental Protection Agency (abamectin 0.41 μg a.i./bee, deltamethrin <0.7 μg a.i./bee, methamidophos <1.4 μg a.i./bee, EPA 2014). These differences are likely the result of methodological differences among the studies, such as the solvent used (acetone in ours and 2-propanol in the EPA database) and period of exposure (24 h in our study and 48 h in the EPA database). Regarding species-related differences, the differential insecticide susceptibility between A. mellifera and M. quadrifasciata observed here may result from differences in the life histories of both bee species. Life histories of different bee species vary considerably, and life story traits (e.g., sociality, body size, target-site sensitivity, and capacity for detoxification by enhanced metabolism) have been linked to differential sensitivity to pesticides (Liu et al. 2005; Hardstone and Scott 2010; Brittain and Potts 2011; Decourtye et al. 2013).
The overall higher insecticide susceptibility of the native species compared with the honey bee emphasizes the limited value of the extrapolation of the results of the toxicity bioassays that compare the latter species to native pollinator species. Several studies have registered sublethal insecticide effects on other bee species despite the fact that these compounds were considered safe in a risk assessment with honey bees (Decourtye et al. 2013); thus, the use of a higher diversity of pollinator species should be considered in such studies.
The Brazilian situation affords an interesting case study due to the economic and ecological importance of the native stingless bee species M. quadrifasciata. This native species is a frequent and important pollinator of native and cultivated species in the country, while the honey bee is an introduced species that does not appear to be a consistent model for assessing pesticide impact in pollinator species (Kevan 1999; Kremen et al. 2002; Tomé et al. 2012). The native species M. quadrifasciata is likely a more suitable model for such studies in Brazil because it is a close relative to native endangered species, notably M. capixaba (Moure and Camargo), which is formally recognized as endangered (Resende et al. 2008; Luz et al. 2011; IUCN 2013). Thus, our findings may contribute to improving the overhaul of the insecticide registration process in Brazil by demonstrating that native bees can be more sensitive to insecticide exposure than honey bees. This information may also support the adoption of suitable programs aiming to conserve and use Neotropical stingless bees in agriculture.
This work was supported by grants from the Arthur Bernardes Foundation (FUNARBE), Minas Gerais State Foundation for Research Aid (FAPEMIG), the National Council of Scientific and Technological Development (CNPq), and the CAPES Foundation.