Evaluating the Impact of Commonly Used Pesticides on Honeybees (Apis mellifera) in North Gonder of Amhara Region, Ethiopia

Global honeybee losses and colony decline are becoming continuous threat to the apicultural industry, as well as, for food security and environmental stability. Although the putative causes are still unclear, extensive exposure of bees to pesticides could be the possible factor for worldwide colony losses. This study was aimed at evaluating the impact of nine commonly used pesticide incidents on adult worker honeybees (A. mellifera) under the laboratory condition, in North Gonder of Amhara region, Ethiopia. Feeding test, contact test, and fumigation tests were carried out for each pesticide following the standard procedures, and each pesticide toxicity was compared to the standard toxic chemical, dimethoate 40% EC (positive control), and to 50% honey solution (negative control). The results revealed that all the tested pesticides caused significant deaths of the experimental bees (P < 0.05) in all the tests when compared to the negative control. Diazinon 60% EC, endosulfan 35% EC, and malathion 50% EC were appeared highly toxic causing 100% mortality of bees, while chlorsulfuron 75% WG killed 90% of the experimental bees as tested via feeding. On the other hand, agro-2, 4-D and its mixture with glycel 41% EC are moderately toxic, and mancozeb 80% WP and glycel 41% EC were slightly toxic to honeybees as compared to the positive control (dimethoate 40% EC). Suddenly, diazinon 60% EC and malathion 50% EC triggered 100% mortality of bees, while endosulfan 35% EC and chlorsulfuron 75% WG caused 63.63% and 90.82% of bee mortality, respectively, when evaluated via contact test. The fumigation test also showed that chlorsulfuron 75% WG, diazinon 60% EC, and endosulfan 35% EC caused 100%, 86.7%, and 65.6% mortality rate of bees. Our result also highlighted that tested LD50 of all pesticide incidents were significantly lower than the manufacturer-based LD50. This shows that local honeybees A. m. jemenetica are extremely sensitive to commonly used agricultural pesticides, which may affect the colony level due to the intensive application of these pesticides in Ethiopia.


Introduction
Honeybees (Apis mellifera) are well known for their commercial products, playing increasing roles in income generation, healthy food, and alternative medicinal values. Tey are not only a key contributor to economic functions but also they are the single most important species pollinator in natural ecosystems across the globe [1]. In Africa, A. mellifera contributes for more livelihood of the community and plays an essential role in pollinating the most of the agricultural crops [2,3]. To this fact, about 50% of the leading global food commodities depend on pollination by honey bees for either fruit formation or seed set [4]. In this case, bees are the most efcient pollinators for most cash crops, stable food crops, vegetables, and fruit trees [5].
However, a large scale dramatic losses and decline of pollinators including honeybees have been reported in several regions of the world resulting severe threat to the apiculture industry and global food security [6][7][8][9]. For example, beekeepers in the United States lost an estimated 50.8% of their managed honey bee colonies only in 2021, which was the highest annual loss on record [10]. Although the extent is diferent, similar trends have been reported in African countries in recent years afecting the selfsustainability of both wild and managed bee populations [11]. Although the putative causes of colony loss are still unclear, the combined efects of climate change, intensive agriculture, pesticides use, pest and pathogens, and biodiversity loss are some risk factors for global honeybee loss [12]. Earlier reports suggested that the extensive exposure of bees to pesticide incidents would possibly be a major factor for honeybee loss and colony decline [13][14][15][16][17]. In Ethiopia, widespread reports indicate that exposure to commonly used agricultural pesticides has been linked to the dramatic honeybee deaths and colony decline than any other factors in the country [18][19][20][21]. Such losses of honeybees have in turn resulted in reduction of honey production as well as crop production, through disrupting pollination services [22].
In general, both managed and wild honeybees are exposed to a wide range of pesticide incidents, which can only be determined through extensive toxicological assessments [23]. In previous studies, a number of pesticide incidents were investigated in several countries of the world [9,24,25]. Te majority of investigations showed that honeybees frequently became exposed to chemical pesticides as a result of their foraging activity. However, some studies suggest that there are three key pathways of poisoning incidents. Te primary incident occurs when forager bees come into direct contact with pesticides that are applied to plants, and the bees rapidly die in the feld [26,27]. A second possible route of pesticide incident happens when forager bees bring contaminated nectar, pollen, and water sources into the hives; thus, the entire colony can be afected by the contaminated material [28,29]. Te third possible exposure of pesticides happens via aerial spray drift [27]. As a result, the measurement of toxic efects of most pesticides has relied largely on the determination of acute toxicity than chronic and sublethal efects due to its rapid appearance of visible symptoms [30]. Acute pesticide tests via ingestion (feeding), contact exposure, or ambient air drifting intake are, therefore, common tests for pesticide incidents [31,32]. Contact exposure and ingestion are well studied routes of contamination that reveal pesticide-specifc efects on honeybee health [33,34]. Nevertheless, exposure of bees to pesticide through air drifting (fumigation test) is thought to be a minor route of pesticide uptake due to volatile nature of some pesticide components [35].
In Ethiopia, intensive application of commonly used agricultural pesticides against pests and weeds control has been largely reported [36][37][38]. Such open feld application of agricultural pesticide incidents has been suspected for most exposure and a fagship poisoning of honeybees in the country. Te majority of farmers in Ethiopia follow an indiscriminate application of pesticides over the open agricultural felds, even during the visiting period of forager bees on the same feld. As a result, beekeepers have been continuously reporting the deaths of honeybees and colony population declining. However, the acute toxicity tests and agricultural pesticide incidents at certain concentration level are not yet studied in Ethiopia.
Terefore, this study was aimed to determine the acute toxicity of nine commonly used pesticide incidents on local honeybees (Apis mellifera jemenitica) under laboratory condition in Chilga district of Northern Gonder, Amhara region, Ethiopia. Understanding the pesticide poisoning incidents through diferent mode of exposures can help to design and implement best management practices in the potential sources of exposure areas.

Pesticide Selection.
Pesticide selection was carried out based on their distribution, wide application, target use, and their market channel in Chilga district of Northern Gonder, Amhara regional state, Ethiopia. Accordingly, nine commonly used pesticides including agro-2,4-D, glycel 41%, diazinon 60% EC, chlorsulfuron 75% WG (or slean 75% WG), mixture of agro-2, 4-D and glycel 41% EC, mancozeb 80% WP, malathion 50% EC, endosulfan 35% EC, and dimethoate 40% EC were purchased from local markets as well as from veterinary drug stores of the Chilga district, Norther Gonder, Amhara region. Te collected pesticides were transported to the regional animal health laboratory and stored at the room temperature (25°C) until the acute toxicity tests were performed.

Bee Samples.
Adult worker honeybees were collected at early in the morning from strong and preassumed healthy colonies based on their activity and internal inspection. Te sampled bees were taken to laboratory using well ventilated plastic jars. Te bees starved for about 2 hrs prior to the commencement of the laboratory experiment in order to induce their pesticide contaminated solution consumption rate.

Laboratory Test.
Acute toxicity of selected nine pesticides (eight widely used and one standard toxic pesticide used as the control) was tested in the laboratory on local honeybees (Apis mellifera jemenitica) via feeding, contact, and vapor tests following the standard laboratory procedure [39,40] (Figure 1). For this purpose, the collected adult worker bees were anesthetized with Co 2 and inserted to well ventilated laboratory cages (size 5.5 × 8.5 × 10 cm), and placed at room temperature (25 ± 2°C) and humidity (60-70%) during study periods. Te mortalities caused by individual pesticides were compared with the positive control (dimethoate 40% EC), negative control (water), and amongst the test pesticides using the following mode of tests.

Feeding Test.
To determine the toxicity efect of each chemical via feeding test, 30 predetermined healthy worker bees were placed in laboratory cages. Ten, the bees were provided with 50% honey solution containing the recommended concentration of 300 μg (logically estimated as 10 μg/bee) of each test pesticide to determine the toxicity efect according to the procedure of Medrzycki et al. [41]. Te recommended concentration of each test pesticide is indicated in Table 1, and each treatment was replicated 3 times. Both the number of dead and injured bees were recorded after 15, 30, and 45 minutes, then after 1, 2, 4, 6, 12, 24, and 48 hrs, and compared with negative control (50% honey solution) and positive standard toxic chemicals (dimethoate 40% EC). Honey solution was replenished for all experimental bees in all test categories whenever required (when they fnished the supplied resource) [42].

Contact Test.
In the mode of contact test, flter papers were immersed in each recommended concentration (Table 1) of test pesticides and allowed to be air dried at room temperature. Te flter papers containing test pesticides were enclosed separately in the lab cage containing 30 worker bees. Toxicity efects of each concentration of test materials were then compared with 0.3% standard chemicals and the control (paper immersed in pure water). Each treatment was replicated three times as described by Gough et al. [42]. Ten, every activity of bees after the application of each test was observed to determine the physiological and behavioral efects of pesticides on experimental bees.

Vapor or Fumigation Test.
For fumigation test, another 30 worker bees were held in laboratory cage and placed over the Petri dish flled with recommended concentration (Table 1) of each pesticide with three replications. Te number of dead and injured bees was recorded in an hour interval for two days. Ten, the death rate of bees was compared with the standard toxic chemical (dimethoate 40% EC) known to kill 100% of bees at concentration level of 0.3% and with nontoxic control (Petri dish flled with water). Similar to feeding test, all the experimental bees in the cages were fed 50% natural honey solution throughout the entire experimental period [42].
Finally, percent of mortality rates caused by each pesticide in each mode of test was corrected by Abbott formula [43] as indicated as follows: % of mortality: Correct mortality (Abbott) � % mortality treatment − % mortality control x 100 100 − Mortality control . (1)

Data Management and Statistical Analysis.
Te variances of laboratory data analyzed using GLM and Tukey's honest signifcant diference (HSD) at 5% level of signifcance were used for mean separation whenever signifcant results were encountered.

Behavioral and Physiological Efect of Pesticides.
In this study, exposure to each pesticide incident appears to impair the behavioral and physiological response of experimental bees in the cages immediately after the exposure ( Figure 1).
We observed that all the experimental bees sufered seriously except for those tested with water control. Typical symptoms for bees sufered due to pesticide exposure include high disturbance, narcotization, hovering sound, and crawling in the bottom wall of the test cages. Tis result highly agrees with the fndings of Tompson [44], who reviewed a wide behavioral efects and potential risks of pesticide incidents on bees following their exposure. Experimental bees in the feeding test were showed high disturbance and narcotization than those in contact and fumigation tests. Fortunately, the bees showed no trophallactic transfer of food from each other as soon as they recognized contamination in the food    Journal of Toxicology solution either to save themselves or loss their cognitive behavior [45,46]. Evidence from recent study also highlighted those pesticide-induced cognitive impairments on olfactory learning, visual learning, and memory of honeybees [7]. Similarly, acute exposure of bees to neonicotinoid induces a series of symptoms that are consistent with hyperresponsive neural impairments [47]. In this case, experimental bees exposed to all the test chemicals showed reduced proboscis extension as compared to the unexposed bees.
Apart from behavioral responses, bees exposed to pesticides showed some observable physiological disruptions. Almasri et al. [48] explained that even mild exposure to pesticides can directly alter the physiological homeostasis of bees and particularly if the individuals exhibit a lack the core microbiota. Meanwhile, such behavioral alteration and physiological disruption caused due to pesticide exposure directly lead to lethal efects on bees at varying time intervals for diferent ages of bees [13,14,48].    (Table 2). However, there was signifcant toxicity diference (P < 0.001) among all the tested chemicals in causing the mortality of bees within the given time intervals. Diazinon 60% EC caused 37.8% and 58.9% at 15 min and 30 min experimental time, respectively, and glycel 41% EC + 2,4-D caused 10% mortality after 2 hrs of exposure to pesticides, which is signifcantly diferent in modes of action and duration (Table 2). Tis implies that the mode of chemical application and exposure time of bees to the pesticide incidents were diferently afecting the bee's lifespan [49]. We also observed that, bees are more signifcantly susceptible to poisoning incidents of the pesticides when ingested the pesticides than via fumigation or body contact tests. In general, the honeybees are exposed to pesticide incidents either through direct contact with pesticides applied to plants during pollen and nectar collection in the feld [26], or through food contamination with the incoming pollen or nectar in the hive [29]. Te signifcant mortality of experimental bees observed in our current investigation could be an indicator for these routs of incident poisoning that has been causing colony losses and decline in Ethiopia.

Feeding Test.
During the feeding test, diazinon 60% EC, endosulfan 35% EC, and malathion 50% EC caused highly poisoning incidents and killed about 100% of experimental bees within shorter test periods (which is in less than an hour), while chlorsulfuron 75% WG killed about 90% of the experimental honeybees ( Figure 1). Tus, these pesticide incidents were comparable to highly toxic standard pesticide (dimethoate 40% EC), but negatively act when compared to the water control. Tis implies that pesticides used by the farmers were fast acting and killed honeybees even before the mid-day if applied in the morning. Tese fndings were partially agreed with the previous fndings of Bezabeh and Gela [50] that stated endosulfan 35% EC and diazinon 60% EC are highly toxic incidents to honeybees of the central highlands during the same laboratory test. Herbicides 2,4-D and glycel 41% EC + 2,4-D killed more than 50% of the experimental honeybees, while fungicide mancozeb 80% WP, and herbicide glycel 41% EC killed 36.7% and 32.22% of the experimental bees, respectively, when ingested with sugar solution. Tese results indicate that 2, 4-D, and glycel 41% EC + 2,4-D, mancozeb 80% WP, and glycel 41% EC caused signifcantly poisoning incidents to the local honeybees as compared to the negative control (P < 0.05). Te mixture of glycel and 2,4-D is highly signifcantly more toxic to honeybees than glycel (P < 0.0001) and is comparable to 2,4-D (P � 0.991) (Figure 2 and Table 3). In the previous study, 2,4-D was reported as nontoxic pesticide among the central highland bees, A. m. bandasii [50], but it was found to be toxic to A. m. jemenitica, while the dose formulation and application method were in the same procedure. Tese poisoning diferences might be due to diferences in geographical races of the bees adapting to specifc stressors. In general, this experimental test suggests that agricultural pesticides might cause severe honeybee deaths during their application period unless necessary precautions are taken.

Contact Toxicity Test.
Contact toxicity analyses of the same nine pesticide incidents listed above were evaluated against the standard highly toxic pesticide (dimethoate 40% EC) and the negative control (nontoxic, water). Laboratory   contact toxicity results revealed that there is highly significant diference between the negative control and pesticide incidents and among each pesticide (P < 0.001). Diazinon 60% EC and malathion 50% EC caused 100% mortality via contact (Figure 3), while endosulfan 35% EC and chlorsulfuron 75% WG killed 63.63% and 90.82% experimental bees, respectively, ( Figure 3) and except endosulfan 35% EC, all were comparably highly toxic like that of standard insecticides, dimethoate 40% EC (P � 0.829 − 1.00). Tis is in agreement with the fndings of Melisie et al. [51] which showed that diazinon 60% EC and malathion 50% EC were highly toxic to honeybees when tested via contact. On the other hand, there was no signifcant contact toxicity difference between the negative control, water, and 2,4-D, glycel 41% EC, and mancozeb 80% WP through contact test (Table 4).

Vapor/Fumigation
Test. Laboratory test of pesticides via vapor or fumigation revealed that all pesticide incidents caused signifcant mortality (P < 0.002) on local honeybees (A. m. jemenitica) as compared to the control treatments (water). Particularly, mortality caused due to chlorsulfuron 75% WG (100%), diazinon 60% EC (86.7%), and endosulfan 35% EC (65.6%) was signifcantly greater than all pesticide incidents tested and was comparable to toxic standards (dimethoate 40% EC) (Figure 4 and Table 5). Tis indicates that these pesticide incidents cause substantial honeybee mortality through vapor, which may be attributed to their fumy properties. Tis result is partiality in agreement with the work of Melisie et al. [51], who indicated that some chemicals including diazinon 60% EC have potential to volatize even at room temperature, and Bezabeh and Gela [50], who showed that diazinon 60% EC caused high mortality on the central highlands honeybees, A, m. bandasii through vapor and ingestion. On the other hand 2,4-D, glycel 41% EC, 2,4-D + glycel 41% EC, mancozeb 80% WP, and malathion 50% EC were less poisonous pesticides than that of toxic standard and they are moderate toxic to honeybees via vapor.

LD 50 for Feeding Test.
In this study, LD 50 of each pesticide was evaluated to support the fndings of the acute toxicity test on experimental bees. Te LD 50 of diazinon 60% EC, endosulfan 35% EC, malathion 50% EC, and chlorsulfuron 75% WG was less than 0.1 μl/bee ( Table 6) indicating that these pesticide incidents were in the standard range of highly toxic substances (LD 50 < 2 μl/ bee) [52]. However, the manufacturer-based LD 50 of these pesticides varies as follows: 0.38 μg/bee for malathion 50% EC, 1.44 mg/kg/duck for diazinon 60% EC, 31-243 mg/kg/ bird, and 30 mg/kg/rat for endosulfan 35% EC (Table 1). Tis shows that commonly used pesticide incidents in Ethiopia were highly poisoning at less concentration than recommended doses and classifed as high toxic pesticides on bees than other animals as recommended by manufacturers. As a result, forager bees are more vulnerable to these poisonous pesticides as compared to other colony members because of their foraging behavior at hotspot areas of pesticide applications. It is expected that some of the foragers may not even return back to hive due to rapid action of these pesticide incidents and thereby causing colony reduction and sever loss of foragers due to these pesticides. In contrast, the LD 50 of 2,4-D, and glycel 41%  Journal of Toxicology EC was tested between 6-8 μl/bee and 5-7 μl/bee, respectively, but manufacturer-based LD 50 for both pesticides is 100 μg/bee (Table 1). Tis also indicates the sensitivity of local honeybees to these pesticides at very less concentration than recommended LD 50 value of manufacturers. While LD 50 of mancozeb 80% WP was ranged between 33-44 μl/bee, which is less than manufacturer LD 50 85.3 μg/bee (Table 1) and, hence, classifed as slightly toxic pesticides (Table 6). In Ethiopia, farmers have used the mixture of 2,4-D, and glycel 41% EC + 2,4-D against herbicides, but caused mild efects on honeybees.
Atkins et al. [53] suggested that this mixture can also be used in the vicinity of bees if dosage, timing, and method of application are in accordance with instructions, but should not be applied directly on bees in the feld or on colonies.

Conclusion
Generally, all the evaluated pesticides which are widely used in the study area (Chilga district) were toxic to local honeybees (A. m. jemenitica) with diferent toxicity levels. Diazinon 60% EC, endosulfan 35% EC, malathion 50% EC, and chlorsulfuron 75% WG poisoning incidents were fast acting and highly toxic to honeybees when tested via feeding and contact. Except malathion 50% EC incident that caused relatively slight toxic efect on honeybees, all other tested pesticide incidents remained highly toxic to honeybees through vapor test. Moreover, 2,4-D and mixture of glycel 41% EC plus 2,4-D are moderately toxic, while mancozeb 80% WP and glycel 41% EC are slightly poisoning incidents to the local honeybees via feeding. Chlorsulfuron 75% WG is herbicide that has been imported illegally to the country and found to be highly toxic to honeybees of the area through all exposure means (feeding, contact, and vapor). In general, this study demonstrated that the tested LD 50 of all pesticide incidents was signifcantly lower than the manufacturerbased LD 50 suggesting that local honeybees A. m. jemenitica are highly sensitive to the commonly used agricultural pesticides in Ethiopia. Terefore, it is an urgent condition to enforce the existing policies, to control, and regulate illegal pesticide marketing, developing policy to exclude the misuse of highly toxic pesticide incidents and set stringent criteria for registration and marketing of less harmful products. As a result, extreme application, illegal import, sale, and distribution of those toxic pesticides should be under strict regulation law enforcement.

Data Availability
All data used to support the fndings of this study are available upon reasonable request from the corresponding author.

Conflicts of Interest
Te authors declare that they have no conficts of interest.