**1. Introduction**

The insect pollination is one of the most important services that sustains biodiversity and food production. Out of the most important commercial crops 84% are insect pollinated [1]. Honeybees (*Apis mellifera* L.) are very important for increasing the quality of fruits and seeds of many wild and cultivated plant species [2,3]. For example, older research in Romania [4] showed the following pollinators' participation in stone tree pollination: honeybees—76.6%; bumblebees—7.6%; flies—3.7%; ants—3.6%; beetles—3.5%; wild bees—2.5%; wasps—0.5%; and other insects—2%.

Due to their complex biology (social life, reproduction, nutrition etc.) and intimate connection to climatic and vegetation conditions, honeybees are a natural biosensor of environmental quality [5–9]. Some of their special behaviors make honeybees a special pollinator: they exclusively feed on nectar and pollen; there is a high number of individuals in a colony, which leads to large quantities of food storage; they have the big ray of forage flight (0–5 km); their "flower fidelity" behavior makes them an efficient pollinator for a certain plant species at a certain moment; they visit many flowers based on the quality and quantity of nectar secretion; they have good orientation, memory and communication regarding food sources; they thermoregulate their nests, which helps overwintering; and they have the possibility to manage and transport their colonies to different crops.

In the context of developing technologies based on electronic sensors combined with the Internet of Things (IoT) and big data storage, honeybees could become even better suited to be equipped and used to monitor di fferent physical and biological aspects of their colonies in order to help beekeepers' management, to avoid high honeybee losses, and to understand the di fferent environmental factors that are causing the declining of their species [9–16]. Thus, from the perspective of managing qualities and their sensitivity to environmental contaminants, the honeybee colony represents an important bio tool to understand and even quantify the impact of di fferent aggressive factors that intersect with its complex social life.

Honeybee depopulation and mortalities in the past two decades have become very important worldwide issues. Though the causes of these issues are quite diverse [17–19], there have been numerous articles on the environmental factors and pesticides. A series of research studies showed that in experimental and field conditions, honeybees' exposure to treated crops registered the following levels of residues: thiacloprid residues were in averages of 75.1 ng/g in pollen and 6.5 ng/g in honey, clothianidin residues were in averages of 9.4 ng/g in pollen and 1.9 ng/g in honey, and imidacloprid residues were in average of 19.7 ng/g in pollen and 6 ng/g in honey [20]. Thiacloprid was also frequently detected in honey samples up to concentrations of 200 ng/g [21], which is the maximum residue level that is accepted for human consumption. In another study [22] the occurrence of pesticides was more frequent in pollen and beeswax, and imidacloprid and fipronil were detected mostly in all matrices. A recent study that was focused on worldwide honey samples [23] showed that 75% of honey samples contained at least one neonicotinoid in quantifiable amounts, the total concentration of the measured neonicotinoids being 1.8 ng/g on average. In another piece of research [24], 97% of neonicotinoids found in pollen came from wildflowers, which grow near treated crops. The presence of more than one pesticide from di fferent categories, in the same sample, is another issue that has been highlighted in much research worldwide that has shown that honeybees are chronically exposed to neonicotinoids and fipronil, and their metabolites in the 1–100 ng/g (ppb) range [20,25,26]. In one study [27], sunflowers and corn tassels contained values of 10 ng/g of imidacloprid in average, which explained why the pollens from these crops were contaminated at levels of a few ng/g. Another study [28] showed that the concentrations of imidacloprid found in sunflower and corn pollens collected in the pollen traps were, respectively, about 1.5 and 4.5 times less than the concentration of imidacloprid found in the same types of pollen that were directly collected from flowers, thus highlighting the potential hazard of neonicotinoids to honeybees through contaminated pollen and nectar. One survey [29] showed that colonies located in a corn-dominated area registered greater colony mortality by 3.51 times more than in corn crop-free locations. In the same study, it was shown that 54% of analyzed samples contained clothianidin, and 31% contained both clothianidin and thiamethoxam.

Studies have shown that these substances a ffect the honeybee organism at nanograms levels (1 ng = <sup>10</sup>−9), with 0.10 ng imidacloprid/bee being the lowest concentration that had detrimental e ffects on honeybees in laboratory conditions [30].

Following the scientific data of numerous studies regarding their toxicity by lethal and sublethal effects, based on European Food Safety Authority scientific reports [31], the European Union (EU) (2018/783/784/785/29.05.2018) banned the use of three neonicotinoids (imidacloprid, clothianidin, and thiamethoxam) in fields. However, a series of countries were approved to use these substances by emergency authorization, Romania being the country which continuously used the forbidden neonicotinoids, at the country level, based on emergency authorizations.

In this context, it is important to mention that Romanian agriculture is one of the main economic sectors that is continuously developing thanks to the favorable geographic and climatic conditions. This situation has led to an increase of land surfaces that are cultivated with industrial crops (sunflower, rape, and corn) and their inputs. In the same time, beekeeping has been a very important sub-domain of agriculture that is favored by natural conditions and stimulated by the organizational national structure under Romanian Beekeepers Association, founded in 1958.

Since Romania's admission into the EU (2007), the developing beekeeping sector has been encouraged by the national beekeeping program and some other agricultural programs that are ruled by European and national legislation. As a result, the number of hives has constantly increased, with the total number of hives being 11.7% of the total EU number with a production of honey of 13% of the total EU honey production in 2018 (these data were published at https://ec.europa.eu/agriculture/ honey/programmes\_en).

If rape and sunflower are very well known in importance for beekeeping, corn crops (*Zea mays*) are not so well highlighted as they are considered a wind pollinated crop. However, even pollen morphology reflects an adaptation to wind pollination, as its nutritional properties make it an important attractant for honeybees and other pollinating insects. Its male flower (the tassel) o ffers large amount of pollen that can be very attractive during the period of sunflower honey flow when crops are nearby. This source of pollen is of grea<sup>t</sup> importance as its flowering period overlaps on that when winter honeybees' generations are reared, so it substantially contributes to the quality of wintering, both by honeybee quality as well as by pollen storage that is consumed in the early stage of brood rearing in the next season. As a consequence, the contamination of corn pollen with neonicotinoids may have a grea<sup>t</sup> negative impact on honeybees [32–37] in the sunflower honey flow period, as a lot of corn fields are closer to the sunflower crop fields—this is a reason why corn crops were included in this study.

Taking into account policy context and certain beekeepers' complaints regarding honeybees' depopulations [38], a research project was funded in October 2017–October 2018 by the Ministry of Agriculture and Rural Development for the first time in Romania. The aim of this project was to establish the realistic-field exposure levels to neonicotinoids in certain areas that are intensively cultivated with oilseed rape, corn and sunflower. To carry out this research, an integrative approach was developed in order to follow up the honeybee colonies and to collect and prepare representative samples in order to evaluate the exposure of honeybees to the neonicotinoids used in intensive crops.

#### **2. Materials and Methods**

This research was carried out in the following phases:


#### *2.1. Field Identification in Di*ff*erent Agricultural Areas*

To implement the protocol and carry out the analyses regarding the exposure of honeybee colonies to neonicotinoids, three fields were selected in di fferent areas of intensive agriculture in the southeastern and eastern parts of Romania, being provided by two di fferent agricultural research stations and one institute that belongs to the Romanian Academy for Agricultural and Forestry Sciences "Gheorghe Ionescu Sisesti". These were located in Neamt county (Statiunea de Cercetare Dezvoltare Agricola—SCDA Secuieni, 46◦5145" N, 26◦4942" E), Arges county (Statiunea de Cercetare Dezvoltare Agricola—SCDA Albota, 44◦4654" N, 24◦4931" E), and Calarasi county (Institutul National de Cercetare Dezvoltare Agricola—INCDA Fundulea, 44◦2710" N, 26◦3055" E). Besides these, one farmers' association was involved in this monitoring study: The Corn Producers' Association (Asociatia Producatorilor de Porumb din Romania—APPR) located in Ialomita county Mihail Kogalniceanu, Tandarei (44◦4039.67" N, 27◦4214.71" E) (Figure 1).

**Figure 1.** Monitored locations by honeybee colonies to field-realistic exposure to neonicotinoids in Romania.

These centers provided different surfaces of land (a minimum of 1 ha and a maximum of 100 ha) that were cultivated with targeted crops (rape, corn and sunflower) and treated with field-prescribed doses of active substances/products by seed dressing, regarding the three neonicotinoids (imidacloprid, clothianidin, and thiamethoxam) that are the subject of interdiction in the European Union. In addition to these locations, a series of samples were collected from two apiaries belonging to the Institute for Beekeeping Research and Development Bucharest, apiaries being located in different areas in southeastern Romania (Baneasa-Bucuresti—44◦2933" N, 26◦0445" E, Buzau—45◦10 N, 26◦49 E), as well as from private beekeepers (Fundulea—44◦2710" N, 26◦3055" E, Otopeni—44◦32 N, 26◦6 E,) (Figure 1), in order to evaluate the presence of neonicotinoids in different areas with intensive agriculture but without any information about the use of phytosanitary treatments.

The experimental honeybee colonies were located near the treated crops (0–100 m), depending on the configuration of the land, at approximatively 10% blooming, in order to attract the honeybees for nectar or pollen collection and initiate the "fidelity flower" behavior.

#### *2.2. Honeybee Colony Preparation and Transportation to Di*ff*erent Envisaged Crop Fields*

In order to monitor the honey/pollen flows and collect representative samples, the experimental honeybee colonies were prepared. In this regard, every location was supplied with two honeybee colonies (normal colonies with queen, brood and food storages), established on 10 Dadant frames each, well covered with honeybees, and equipped with entrance type pollen collectors and foundation frames for honey collection in order to let honeybees build combs in the honey flow conditions to decrease the risk of contaminating the collected honey by older combs.

Out of these colonies, one honeybee colony per each location was equipped with an electronic hive (Simbee®, http://www.simbee.ro/) that consisted of a special module of sensors and data collectors: two ambient temperature and humidity sensors, an internal sensor for honeybee colony temperature, and a weight sensor scale.

The monitoring hives transmitted the collected data to its database every 10 min, thus registering the activity of bee colonies regarding the monitoring parameters and helping to understand the existence or lack of honey flows in the flight area, as well as suspicions about a possible decline of population and its development status that are connected with phytosanitary treatments.

#### *2.3. Sample Collection and Specific Preparation*

#### 2.3.1. The Honeybee Samples

These samples were collected from dying or live honeybees, depending on the encountered situation. The dying honeybees were collected out from the front of hive and, when faced a lack of dead or dying honeybees, we collected live honeybees (foragers) from the entrance, following the honey or pollen flow, by using a car aspirator. The samples were immediately confined to a bag and put in a car freezer.

As during the experiments, acute and lethal effects on honeybees were not directly observed in most of the cases, the honeybee sampling consisted of the collection of live forage honeybees from the entrance of the hives in order to increase the probability of finding neonicotinoids in honeybees that carry freshly contaminated nectar or pollen.

#### 2.3.2. The Honey and Pollen Samples

After 7–10 days from the beginning of the honey flow, honey samples were collected from specially prepared and introduced frames in the monitored hives that were preserved in refrigerators.

The collected honeybee pollens were taken out from collectors and preserved in specific low temperature conditions (between −10 and 4 ◦C) daily, depending on local situation, until their mono-floral analyses and special preparing samples, which was the case for samples sent to the French Agency for Food, Environmental and Occupational Health and Safety (ANSES) laboratory.

In order to identify the source of different contaminants by using the honeybee as a sampler, one problem was represented by the big ray of forage flight from the hive (0–5 km), which covers a big surface of land and seldom obtains multi-floral products, e.g., honey, pollen, and beebread. In the correct monitoring of a specific crop, the problem lies in collecting its specific mono-floral samples, as this has a big impact on residue data identification and interpretation.

Thus, to understand the floral componence and to establish the mono-floral honeys from the targeted field, the samples collected out from the rape and sunflower crops were analyzed in the framework of the chemistry laboratory of the Institute for Beekeeping Research and Development, based on specific standardized methods of melissopalinology used in the evaluation of honey types (Table 1). These analyses are very important in the context of the neonicotinoid residues analysis because they confirm whether the samples are sufficiently relevant for the purpose of the study.

Due a lack of mono-floral envisaged samples because of climatic conditions (drought or heavy rains specific to the 2018 season) in the present study, we also used honey collected during the studied crop flowering that was classified as multi-floral honeys but also contained the targeted honey in different percentages.

Regarding the mono-floral pollen samples and taking into account the variability of pollens usually collected by honeybees, most of the pollens collected by specific entrance collectors in the rape and sunflower period were multi-floral. To have relevant, envisaged mono-floral pollen samples for neonicotinoid analyses, based on the minimum required quantities (when possible), we manually selected the pollen pellets of rape, corn and sunflower (Figure 2a,b) in the laboratory conditions based on the pellet aspect and color, with the selection being randomly confirmed by microscopy based on pollen characteristics (Figure 3a–c).

Pollen mono-floral selection, a time consuming activity, was possible only when the minimum sample quantity requested by laboratory was low. The minimum quantities of the samples requested by the two laboratories were a minimum of 10 g at the ANSES laboratory and a minimum of 250 g at Quality Services International (QSI) laboratory. The minimum quantity of 10 grams (e.g., ANSES laboratory) permitted a better approach in the sample preparation, thus providing a better analysis of the pollen origin regarding the plant species. The process of collection and selection of representative samples was a key stage in the neonicotinoid identification and quantification.


**Table 1.** The obtained results on different types of honey samples, following the mellissopalinological analyses.

**Figure 2.** Example of honeybee multi-floral pollen samples collected by honeybees in the two studied periods. The rape, sunflower and corn pollens pellets were selected in the laboratory when preparing the samples to be analyzed. (**a**) A pollen sample collected during the oilseed rape (*Brassica napus*) blooming. One can notice an amount of approximately 50% oilseed rape pollen, (indicated by red circles). (**b**) A pollen sample collected during the sunflower (*Helianthus annuus*) and corn (*Zea mays*) blooming. One can notice an amount of approximately 60% sunflower pollen (orange) versus 40% corn pollen (yellow). Photos© Institute for Beekeeping Research and Development, Bucharest.

Following these selection steps, 50 samples were prepared and sent to the laboratories, as follows: honeybees—10 samples; honey—15 samples; and pollen—25 samples.

**Figure 3.** Microscopic view of the pollen grains used to confirm the studied selected samples. (**a**) The oilseed rape (*Brassica napus*) pollen had a ~24 μm grain with a round tricolpate form and grooves at the rounded tip, covered with a membrane, sometimes with small granular residues; the exine had a finely cross-linked surface. (**b**) The corn (*Zea mays*) pollen grain was a ~99 μm granule with a single porous opening, the exine and the intine being thin membranes and the cytoplasm appearing granular with numerous small starch formations. (**c**) The sunflower (*Helianthus annuus*) pollen grain was ~35 μm and was a trickled grain, having a surface covered in thin and long thorns that protruded to the pores. The used equipment: Olympus, Quick Photo camera 3.1., ×200: Photos © Institute for Beekeeping Research and Development, Bucharest.

#### *2.4. Sample Preservation, Codification, Packing and Sending to the Accredited Laboratories According to Specific Requirements*

After the collection and transportation to the central laboratory of the Institute for Beekeeping Research and Development, all the samples were preserved at −18 ◦C.

For identification, samples were coded with a unique code, e.g., R-P-F-2 = Rape-Pollen-Fundulea-2nd sample (R = rape; FS = sunflower; P = corn; M= honey; P= pollen; and A= honeybees). The samples, which were prepared according to the specific requirements of analyzing laboratories, were packed in special containers and sent to the following two European accredited laboratories: thirty samples were sent to an EU Reference laboratory—French Agency for Food, Environmental and Occupational Health and Safety (ANSES), France, and 20 samples were sent to Quality Services International (QSI), Germany. The shipping was performed in special freezing packing for biological material.

The two laboratories were chosen in order to diversify and better understand the requirements of the accredited laboratories regarding the preparation of samples for analyses. Generally, the samples were distributed between the two laboratories depending on the collected quantity correlated with melissopalinological analysis.

The results of neonicotinoid analyses performed by the accredited laboratories are presented in Tables 2 and 3, together with their level of detection and quantification on five neonicotinoids—acetamiprid, clothianidin, thiamethoxam, imidacloprid and thiacloprid. The analyses were done with the liquid chromatography method coupled with tandem mass spectrometry (LC-MS/MS) at both laboratories.


**Table 2.** The results on the neonicotinoid residues found in samples collected from different crops and locations in Romania (2018).

*Diversity* **2020**, *12*, 24


**Table 3.** The limit of detection and quantification of the involved laboratories, as well as the maximum residue limit in conformity with European legislation.

3 LOD = limit of detection; 4 LOQ = limit of quantification; 5 MRL= the maximum residue level permitted for humanconsumptionaccordingtoEuropeanlegislation.
