Results show that T. angustula is sensitive to all doses of glyphosate tested. The mortality observed after 24 and 48 hours at each dose in the lethality test showed that deaths increase with increasing pesticide dose (Fig. 2A and 2B). No bee in the control group died during the analyzed period.
Similar results with glyphosate were observed in other studies using T. angustula and A. mellifera (Toledo and Guillén 2014; Lunardi 2018, respectively). Yet, at the 50µg/bee dose, the highest tested, the number of deaths was lower than that recorded at the 25, 10, and 5µg/bee doses, showing that the effects cannot always be linear (Suchail et al. 2000), not only as a function of the dose but also of the bee's metabolism.
Observational LD50 was 0.25µg/bee during 24 hours observation. By probit, LD50 was 0.205µg/bee, with a 95% confidence interval from 0.09µg/bee to 0.47µg/bee. In 48 hours, the observational LD50 was 0.025µg/bee and by probit, 0.015µg/bee, with a 95% confidence interval ranging from 0.005µg/bee to 0.04µg/bee. As classified by Atkins et al. (1981), the LD50 obtained indicates that glyphosate is a highly toxic product for T. angustula, which can cause great damage to bee populations if used close to them.
The LD50 of the present work differed from the one described by Lunardi (2018), of 273.93µg/bee by ingestion and 255.73µg/bee by contact with Apis mellifera in 24 hours. Different genetic susceptibility of bee species may explain this difference, as differences between the techniques used, besides environmental conditions (Moraes et al. 2000; Pereira 2010).
Glyphosate had an LD50 similar to insecticides applied orally and topically on A. mellifera, such as Chlorpyrifos (LD50 = 0.25µg/bee, oral) (Suchail et al. 2000) and Malathion (0.27µg/bee, topical) (Smart and Stevenson 1982). These demonstrate that insecticides and herbicides can have the same lethal effect on bees and that hives close to agricultural environments can be affected by those agrochemicals.
If a bee ingests 5µL of glyphosate at a concentration of 0.0026gL-1, as found in nature and used in this research, it will ingest 0.013µg/bee of glyphosate. A value like our probit 48h-DL50 of 0.015µg/bee, lying within the confidence interval. So, bee mortality may be occurring in nature.
At the beginning of the tests, bees were in the same condition as shown when comparing the pre-ingestion tests (hour 0) at all doses, with no statistical difference between locomotion times or behavioral changes.
Glyphosate increased the locomotion time of bees at the doses tested. Comparisons of the times for bees to travel 50cm before the ingestion of the solutions (0h) did not reveal a significant difference (H = 10.6400, GL = 6, p = 0.1002) between the groups of bees that would receive glyphosate and the control group (Fig. 3A).
One hour after glyphosate ingestion, there was a significant difference (H = 136.0303, GL = 6, p < 0.0001) between the tested groups and the control group, except at the dose of 0.0025µg/bee (Fig. 3B). After 4 hours and 24 hours of ingestion of the solutions, there was a significant difference (H = 140.7984, GL = 6, p < 0.0001; H = 122.8370, GL = 6, p < 0.0001, respectively) between the tested groups and the control group.
Similar effects on locomotion were found with glyphosate in Melipona capixaba at a concentration of 17.5gL-1, decreasing the survival of individuals and impairing flight (Gomes 2017). In A. mellifera, concentrations of 100gL-1 caused 100% mortality (Serpa, 2017), while 0.01gL-1 in the field, altered hive return time, interfering with the species’ cognition (Balbuena et al. 2015).
For other products, there are also reports of effects like those caused by glyphosate in the present work. For Melipona quadrifasciata (Tomé et al. 2015), Melipona scutellaris (Costa et al. 2015)d angustula (Quiroga-Murcia et al. 2017; Jacob et al. 2019), neonicotinoid insecticides changed behaviors (Quiroga-Murcia et al. 2017), impaired motor functions, and locomotion, and changed average speed and distance traveled (Jacob et al. 2019).
The results differed from those found by Souza (2021) in Scaptotrigona aff. xanthotricha, in which glyphosate, in a concentration of 1.375L/100L of water, reduced the running time of the bees. The increase in locomotion time observed in this work can also occur in the field, decreasing bee speed and foraging efficiency, adversely affecting pollination services (Jacob et al. 2009).
Locomotion time increase had higher statistical significance at higher doses, indicating that higher dosage causes more intense effects and longer travel time, especially near 24h. Similar to the results of Lunardi (2018) testing glyphosate in A. mellifera. Those findings may indicate a slow metabolization of glyphosate by bees with a higher risk of intoxication (Yue et al. 2018).
Control solutions at 0, 1, 4 and 24h had no statistical significance (H = 6.1812, GL = 3, p = 0.1031) (Fig. 4A). There was a statistical significance between different times for each dose. Doses of 0.0025µg/bee, 0.013µg/bee, 0.025µg/bee, 0.038µg/bee and 0.25µg/bee, showed a statistical significance (H = 38.7700, GL = 3, p < 0, 0001; H = 58.2975, GL = 3, p < 0.0001; H = 54.3401, GL = 3, p < 0.0001; H = 71.7682, GL = 3, p < 0.0001 and H = 59.3940, GL = 3, p < 0.0001, respectively) between 0h (before receiving the dose) and the other hours (1, 4, and 24h after receiving the dosage. The difference was greater at 4 and 24h after application (Fig. 4B, C, D, E and F).
Hundred fifty bees that received glyphosate, but 67 did not complete the route (44.6%), of which 19 returned to the light (12.6%), 21 stayed in the lane (14%) and 27 died (18%) (Table 1).
Table 1
Bees that did not complete the course. N = 30 for each dose at each time.
Time Course | Dose (µg/bee) | Did not complete the route |
Returned to the light | Stayed in the lane | Died | Total |
1h | 0.25 | 2 (6.6%) | 1 (3.3%) | - | 3 (10%) |
4h | 0.013 | - | 2 (6.6%) | - | 2 (6.6%) |
| 0.025 | 1 (3.3%) | 1 (3.3%) | - | 2 (6.6%) |
| 0.038 | 3 (10%) | 2 (6.6%) | - | 5 (16.6%) |
| 0.25 | 5 (16.6%) | 6 (20%) | 1 (3.3%) | 12 (40%) |
24h | 0.0025 | 1 (3.3%) | 1 (3.3%) | 1 (3.3%) | 3 (10%) |
| 0.013 | 2 (6.6%) | 2 (6.6%) | 2 (6.6%) | 6 (20%) |
| 0.025 | 1 (3.3%) | 2 (6.6%) | 2 (6.6%) | 5 (16.6%) |
| 0.038 | 3 (10%) | 3 (10%) | 5 (16.6%) | 11 (36.6%) |
| 0.25 | 1 (3.3%) | 1 (3.3%) | 16 (53%) | 18 (60%) |
Total | | 19 | 21 | 27 | 67 |
Source: Own elaboration. |
In addition, among the 150 bees, 34 (22.6%) presented excessive self-cleaning behavior and 42 (28%) presented tremors, and 5 performed both. Of the 34 who showed excessive self-cleaning, 29 completed the route, and among the 42 who had tremors, 33 finished the course (Table 2).
Table 2
Excessive self-cleaning behavior and tremors observed during the course. N = 30 for each concentration at each time.
Time course | Dose (µg/bee) | Did not complete the route | Completed the route |
Excessive self-cleaning | Tremors | Excessive self-cleaning | Tremors |
1h | 0.25 | - | 1 | 6 | 3 |
4h | 0.0025 | - | - | - | 1 |
| 0.013 | - | - | 1 | - |
| 0.025 | - | 2 | 1 | 2 |
| 0.038 | 3 | 3 | 4 | 5 |
| 0.25 | - | - | 4 | 5 |
24h | 0.0025 | 1 | - | - | 1 |
| 0.013 | 1 | 1 | 4 | 3 |
| 0.025 | - | - | 3 | 4 |
| 0.038 | - | 2 | 3 | 4 |
| 0.25 | - | - | 3 | 5 |
Total | | 5 | 9 | 29 | 33 |
Source: Own elaboration. |
T-test and Wilcoxon testes showed statistical significance before and after glyphosate ingestion on all behaviors analyzed (excessive self-cleaning, turning back to light, tremors and standing still) (Fig. 5), indicating that the pesticide causes the observed changes. Analysis of the control group and before ingestion (0h) tests did not show those behaviors, but they were present at 0.25µg/bee at the one-hour observation.
In the natural environment, these effects affect the ability of adults to perform their functions, such as foraging, and make entry into the hive difficult, as they are attacked and rejected by guard bees, damaging the colony (Johansen and Mayer 1990; Freitas and Pinheiro 2010).
In nature, bees use the sun as a reference for orientation (Bomfim et al. 2017), and in a dark environment, they present positive phototaxis (Lambin et al. 2001). Comparing the lane light with the sunlight, the movement against the light indicates a change in orientation behavior and cognition, which can mean an inability to return to the colony in the wild.
Studies with deltamethrin (insecticide) also caused disorientation. Sublethal doses affected flight muscles and coordination and made it difficult for A. mellifera to return to the hive, suggesting a loss of integration of the visual pattern concerning orientation by the sun (Vandame et al. 1995).
A. mellifera bees exposed to permethrin (insecticide) lost their ability to orient and presented severe behavioral disturbances, such as excessive self-cleaning, similar to the observed in the present study. These disturbances affect foraging ability and impact colony development and survival (Cox and Wilson 1984).
The neonicotinoids, carbamates, pyrethroids, and organophosphates alter sodium transport and compete with acetylcholine. These collapse the nervous system causing poisoning symptoms such as hyperexcitation, convulsions, hypersensitivity, tremors, uncoordinated movements, and paralysis (Soderlund et al. 2002; Faria 2009; Silva et al. 2016). All those combined affect bees’ mobility.
Glyphosate action mechanism in bees is not understood yet, but T. angustula tremors and inability to move may indicate a neurotoxic action, generating a loss of motor functions, like the ones caused by insecticides. As such, Chaves (2020) showed that glyphosate was toxic affecting T. angustula antennations and trophallaxis frequency.
Mortality and all observed sublethal effects are detrimental to bee survival, and those sublethal effects are a problem since experiments that confer licenses for pesticide use only analyze acute toxicity to bees (Cham et al. 2020). Also, current official protocols used to determine pesticide toxicity consider only its effects on A. mellifera (OECD 1998; Thompson 2003). Researchers have described more than 400 species of stingless bees, and it is necessary to standardize the studies using representatives of the native fauna as model organisms.
It is necessary to minimize the harmful effects and reduce the use of pesticides in Brazilian agricultural ecosystems, demonstrating the need to change regulatory policies for the registration, release, and monitoring of pesticides. It is necessary to consider the safety of pollinators concerning lethal effects, but also long-term sublethal toxic ones, considering herbicides and not just insecticides.
Colonies' honey quality can change when contaminated nectar transfer occurs onto the hive and by trophallaxis. Honey quality affects all colony members, altering oviposition and brood rearing, decreasing hive vigor, and leading to death (Free 1980; Sánchez-Bayo et al. 2016).
In A. mellifera, glyphosate also causes metabolic changes (20 mg/bee) (Helmer et al. 2015), reduces sucrose sensitivity and memory, and impairs learning and the odor-reward association (5.0mgL-1e 2.5mgL-1) (Herbert et al. 2014). It also reduces hive population, causes frequent absence of queens, and death of pupae promotes changes in the hypopharyngeal glands cells responsible for royal jelly production (0.0075mgL-1) (Faita et al. 2018), and affects the composition of the beneficial intestinal microbiota of bees (10mgL-1 and 5mgL-1), increasing the susceptibility to problems by pathogens (Motta et al. 2018).
All these behavioral and metabolic changes can lead to hive abandonment in response to the disturbance (Almeida 2008) in an attempt to survive. However, the relationship between the abandonment of nests and pesticides is unknown, and there are no guarantees that bees will restore a new nest, which may cause colony losses.
This study analyzed acute effects in adults. Further investigations are needed to understand if bees can recover after 48 hours and at different development stages to determine larval mortality rate as proposed by Zhu et al. (2014) and glyphosate's long-term effects on colony behavior and survival.
In conclusion, the results suggest that glyphosate can negatively affect the pollination activity of T. angustula. Toxicity and sublethal effects evaluation of pesticides on bees contributes to a better understanding of the harmful effects on hives and to adopt strategies to reduce intoxication.