We used irregular shaped fluorescent polystyrene (PS) particles labelled with rhodamine-b (λ_max = 544 nm) for feeding trials, as PS is one of the most abundant MPs found in the environment (Wagner et al. 2014). Fluorescent plastic granules (Magic Pyramid Brücher & Partner KG, Frechen, Germany) were milled (centrifugal mill ZM 200, RETSCH GmbH, Haan, Germany; rotor: 24Z; sieve: distance sieve 120 µm) and subsequently sieved to achieve the two particle size classes we worked with. In the first size class 50% of the particles had a diameter (d₅₀) smaller than 40.5 µm (d₁₀ = 26.86 µm, d₉₀ = 127.1 µm; Fig. A1; thereafter called “fine MP”) and the second a d₅₀ of 163 µm (d₁₀ = 95.74 µm, d₉₀ = 303.6 µm, Fig. A2; thereafter called “coarse MP”). The sizes were determined by a Microtrac Sync particle analyser (Microtrac RETSCH GmbH, Haan, Germany).

For the feeding trials with bumblebees, we additionally used tire wear particles (TWP) from tires of a passenger car, provided by Continental AG with a known composition of ingredients (see Tab. A3), representing a real-life tire tread material. Tire wear particles were generated by Continental AG using a tire buffing machine. In the central service project Z01 of the CRC 1357 Microplastics at the University of Bayreuth particles were sieved to obtain defined size fractions. The fractionation of the tire wear particles was carried out by a Hosokawa Alpine air jet sieve e200 LS (Hosokawa Alpine AG, Augsburg, Germany) using two different sieves of mesh sizes of 125µm and 200µm. Particles were collected via a cyclone connected to a vacuum cleaner into glass bottles. We obtained three size fractions with particle sizes bigger 200µm, 125–200µm and less than 125µm. For this study the size fraction 125–200 µm was used with a d₅₀ of 147.9 µm (d₁₀ = 94.41µm, d₉₀ = 214.6 µm, Fig. A4), determined by a Microtrac Sync particle analyser (Microtrac RETSCH GmbH, Haan, Germany). The size of the coarse MP particles and the size of the TWP were chosen to largely overlap. Thus, we could easily see if oral ingestion of particles in this size range would regularly occur in this set up, since in contrast to the coarse MP used, the TWP are not fluorescence labelled. The size of the fine MP particles was chosen in the size range of pollen B. terrestris naturally ingests (Velthuis and Cobb 1991; Free 1993; Rasmont et al. 2008; Kämper et al. 2016; Raine et al. 2022). Therefore, this treatment would help to observe the pathways of ingested MP particles without the possibility of size-obstacles.

The ants and bumblebees were kept in a climate chamber at a constant temperature of 26°C and 70% humidity under an inverted 12:12 h dark:light cycle. The Camponotus japonicus colony (ANTSTORE, Berlin, Germany) was fed twice per week with honey water (2:1 ratio of water and honey) and cockroaches (Blaptica dubia, in-house breeding) ad libitum. Bombus terrestris colonies were ordered from Biobest (Westerlo, Belgium). The colonies were kept in their delivery boxes, and each was fed three times per week with ~ 10 grams of pollen (Imkerpur, Osnabrück, Germany) and sugar water (1:1 ratio of water to Apiinvert (Südzucker AG, Mannheim, Germany)) ad libitum.

In our proof-of-principle study, the focus was not on detecting effects of MP/TWP-exposure in terrestrial insects, but on observing the potential transfer of particles within colonies of social insects. Consequently, the concentrations we used were significantly higher than MP concentrations that currently occur in natural environments. For instance, with 0.4% w/v of MP/TWP particles in the feeding suspensions, we provided the bumblebees with concentrations approximately a thousandfold higher than MP-concentrations, that can currently realistically be found in nature (e.g. 4.5 mg per kg in dry weight of agricultural soils, Büks and Kaupenjohann 2020). Since ants possess the IPB as a powerful filtering device, we chose even higher concentrations of MP (2% w/v). We chose these concentrations to facilitate the recovery of the particles in the insect body. Currently these MP concentrations significantly exceed concentrations that are considered environmentally relevant. However, MP accumulation in the environment is predicted to increase exponentially (e.g. Meizoso-Regueira et al. 2024), and the concentrations used in our study could be realistically reached over the course of the next century (Meizoso-Regueira et al. 2024: models predict 0.67% w/w of MP in soils by 2122).

For the MP-exposure experiment, we used four plastic boxes (PP, 20 cm x 14.5 cm x 9 cm) where the bottom was filled with about 1 cm of plaster. In each box we placed a falcon tube, filled with 15 ml deionized water and a cotton ball pushed to the 15 ml mark to provide constant moisture. Afterwards we carefully transferred 60 minor workers and 35 larvae into each box to establish microcolonies. Two microcolonies were randomly assigned as a control (sugar water only) and two for the fine MP exposure (sugar water mixed with 2% w/v fine MP and 0.02% v/v Tween-80: SIGMA-ALDRICH CHEMIE GmbH, Taufkirchen, Germany). The use of controls in our proof-of-principle study was important to check for autofluorescence of body structures of the insects and ensure clear differentiation of MP particles from such. The solutions for both treatments were freshly prepared each week and stored at 4°C. The solutions were vortexed for 10 seconds daily, after which the ants were fed with 400 µl of their respective solution. For C. japonicus we did not use suspensions containing coarse MP since this would likely not pass the proventriculus, a valve that mechanically restricts the passage of fluids from the crop into the midgut in adult ants (Eisner and Wilson 1952).

We continued the feeding trials until at least 10 larvae pupated per box. Subsequently, the workers and remaining larvae were euthanized with ethyl acetate vapour and stored in 4% paraformaldehyde (PFA) in PBS (v/v; pH = 7.4; MORPHISTO GmbH, Offenbach am Rhein, Germany) at 4°C until further analysis. The pupae were kept in a petri dish in the climate chamber under the above-mentioned conditions until they reached the P7 stage (abdominal and antennal pigmentation has finished; Ishii et al. 2005), so we could observe which tissue potentially contains MP. Afterwards, they too were euthanized with ethyl acetate and stored in 4% v/v paraformaldehyde at 4°C until further analysis.

For the exposure to MP, we provided the queenright B. terrestris colonies with treatment solutions/suspensions, in addition to their ad libitum sugar water supply. This was necessary, since the particles cannot be provided through the wick, through which the sugar water is provided (Fig. A5, Fig. A6). We used 15 queenright B. terrestris colonies to obtain 3 colony-replicates per treatment. The colonies were assigned randomly to one of five treatments: control, solvent control, fine MP, coarse MP and TWP. The control was supplied with pure sugar water, the solvent control with sugar water containing 0.02% v/v Tween 80, the fine MP treatment with sugar water, 0.02% v/v Tween 80, and 0.4% w/v fine MP, the coarse MP treatment with sugar water, 0.02% v/v Tween 80, and 0.4% w/v coarse MP, and the TWP treatment with sugar water, 0.02% v/v Tween 80 and 0.4% w/v TWP. For the provision of the treatment solutions/suspensions, we modified the feeding areas of the colonies and placed the treatment solutions underneath the bottom grid. Thus, we ensured that the workers could reach the contaminated food with their mouthparts only, in order to prevent bumblebees from spreading the particles throughout the colony via body contact (for further details see pictures in Fig. A5 and schematic set-up with description in Fig. A6).

To prevent growth of mould or bacteria the solutions in the syringes were changed three times per week and the sugar water tanks were refreshed once a week. Parallel to syringe-changing, the colonies were fed with ~ 10 grams of pollen (organic flower pollen: DE-ÖKO-037, Imper Pur; Osnabrück, Germany).

Bumblebee colonies were fed over a period of four weeks, to ensure the development of pupae, and the emergence of imagines, which both have potentially been fed with MP by their nestmates as larvae. After four weeks the colonies were anesthetized with CO₂ by placing the colony boxes on dry ice in a closed styrofoam container for several minutes. Subsequently, queens, larvae, pupae, imagines, honey- and wax-samples were collected from the colonies. The colonies were then euthanized in the freezer at -20°C overnight.

From each microcolony, 20 randomly selected ant workers and all larvae and pupae were observed using a Leica M205 FCA fluorescence stereo microscope (Leica Microsystems GmbH, Wetzlar, Germany) equipped with a Leica DBC6200 camera and the Leica Application Suite X software (version 3.8.2.27713) to detect MP on their body surface. If MP was found on the individual, it was carefully submerged in acetone for 10 seconds to dissolve the PS and eliminate the fluorescence. The procedure was repeated until we observed no more fluorescence on the surface. This ensured that any fluorescent particles detected in the following steps could only have come from MP that was present within the body.

We checked the digestive system of workers, including the IBP, and the digestive system of larvae, that do not possess an IBP, and the entire body of pupae for the presence of MP. In worker ants we analysed the IBP and the rest of the digestive system that lies posterior of the IBP separately from each other (described as ‘IBP’ and ‘digestive system following the IBP’ hereafter). For our investigations we used two different methods: 50% of the individuals investigated per microcolony were dissected under a stereo microscope while the rest was subjected to a tissue clearing protocol described by Ritschar et al. (2022). Tissue clearing enables in situ detection of MP particles without prior dissection.

For the tissue clearing process of the ants an adaptation of the CUBIC (Clear Unobstructed, Brain Imaging, Cocktails and computational analysis) protocol (Susaki et al. 2015) was used with the following procedure: the individuals coming from the 4% PFA in PBS were rinsed in PBS (pH = 7.4) three times for five minutes each. Afterwards, they were placed in embedding cassettes (Simport Scientific Inc., Boleil, Canada) in sample containers (Faust Lab Science GmbH, Klettgau, Germany) with each separate container holding one live stage (e.g., adult, larvae, or pupae) of one treatment. The loaded containers were then filled with 0.03% H₂O₂ in EtOH and placed on a shaker at 37°C with renewal of the solution every other day for depigmentation. When no further bubble formation occurred, and hence no further bleaching of the individuals could be observed in the current H₂O₂ solution, the concentration was first increased to 3% and subsequently, when also here no further bleaching was visible, to 5%. The slow increase of the H₂O₂ had to be performed to avoid excessive oxygen production and hence to reduce air bubble entrapment in the specimens as good as possible. When the depigmentation procedure was completed, removal of the lipids followed. Therefore, the ants were placed in ½ CUBIC-1 (1:1 PBS and CUBIC-1) overnight (also on a shaker at 37°C). For 500 g of the CUBIC-1 solution 125 g urea (Grüssing GmbH, Filsum, Germany) and 156 g of 80 wt% quadrol (N,N,N′,N′-Tetrakis(2-hydroxypropyl)ethylenediamine, Sigma-Aldrich, Merck KGaA, Darmstadt, Germany, dissolved in milli-Q H₂O were dissolved in 144 ml milli-Q H₂O on a hot plate at 80°C. After all ingredients were dissolved, 75 g Triton-X-100 (Thermo Scientific, Life Technologies GmbH, Darmstadt, Germany) were added under continuous stirring. After the ½ CUBIC-1 solution, the samples were submerged in CUBIC-1 until the ants were transparent. For adjustment of the refractive index, the samples were placed in ½ CUBIC-2 (1:1 PBS and CUBIC-2) overnight on a shaker and at room temperature. For 500 g of the CUBIC-2 solution 125 g urea and 250 g sucrose were dissolved in 75 g of milli-Q H₂O on a hot plate at 60–65 ^oC. Then the solution was cooled to room temperature and 50 g of triethanolamine (SIGMA-ALDRICH CHEMIE GmbH, Taufkirchen, Germany) were added under continuous stirring. After the ½ CUBIC-2 solution the specimens were placed in CUBIC-2 (also on a shaker at room temperature). Then they were screened for fluorescent MP using a Leica M205 FCA fluorescence stereomicroscope (Leica Microsystems GmbH, Wetzlar, Germany) equipped with a Leica DBC6200 camera and the Leica Application Suite X software (version 3.9.0.28093).

Similarly, for analysing B. terrestris for the presence of MP, we formed two sets of individuals, one set for dissection and one set for tissue clearing. For dissection, 15 workers, 15 larvae and 9 pupae were randomly selected per colony and stored frozen until examination. All specimens were checked for MP on the body surface using the Leica M205 FCA fluorescence stereo microscope before dissection. In preliminary tests we found that especially the larvae and pupae were very brittle once dipped into acetone, which complicated dissection. Therefore, if MP was found on the body surface, we instead removed the MP with 70% EtOH until the body surface was MP-free. Only then were the digestive systems dissected from workers and larvae, while the pupae were crushed and analysed for MP. The other set of individuals (one worker, one larva, and five pupae per colony, from 15 colonies) was prepared for tissue clearing by conservation in 4% PFA (in PBS) following the protocol of Ritschar et al. (2022) for B. terrestris. Workers, furthermore, were previously shaved. If specimens were damaged during handling, they were excluded from the analysis.

The tissue clearing process of the bumblebees was similar to the clearing of the ants with minor changes only. At first, the specimens deriving from 4% PFA were rinsed in 1x PBS thrice for five minutes and then placed in syringes for the subsequent steps. Therefore, adult and larvae individuals were placed separately in 20 ml and 10 ml syringes (Omnifix and Injekt syringes, B. Braun SE, Melsungen, Germany) respectively, whereas always five pupae (deriving from the same colony) were pooled in a 20 ml syringe for further processing. After the washing in 1x PBS, the samples were also depigmented, using the same rising H₂O₂ concentrations as for the ants with exchange of the solution every other day. When the depigmentation procedure was completed, the adults and pupae were transferred from the syringes to centrifuge tubes (15 ml and 50 ml) for further handling. Then the lipids were also removed by placing the bumblebees in ½ CUBIC-1 overnight (also on a shaker at 37°C) and subsequently in CUBIC-1 until they were transparent. As the depigmentation and clearing steps took longer for the bumblebees than for the ants, the bumblebee individuals were more fragile. Therefore, to avoid tissue damage, they did not go through the ½ CUBIC-2 and CUBIC-2 steps but instead were immediately screened for fluorescent MP as described for the ants.

All statistical analyses were conducted using R 4.4.1 (R Core Team 2024). The barplots were created and arranged using the packages ggplot2 (ver. 3.5.2; Wickham 2016) and ggpubr (ver. 0.6.0; Kassambara 2023) respectively. To detect statistically significant differences in MP/TWP particle presence among the treatments, we conducted Fisher’s Exact Test (fisher.test()) for our ant data and Pearson's Chi-squared Test (chisq.test()) for our bumblebee data.

The feeding trial took four and nine days, respectively, for at least 10 larvae to pupate in the two control microcolonies, and seven and 15 days, respectively, for the MP treated microcolonies. In the latter case, we prematurely ended the feeding trial after 15 days with only 8 pupae, since the workers started to feed on their own larvae. We observed that workers from all microcolonies were feeding on their supplied sugar water and performed trophallaxis with their nestmates. Five control workers each had a single fluorescent particle in their IBP, probably due to cross contamination. However, we found significantly more workers of the fine MP treatment with fluorescent particles in their IBP (40 out of 40 workers) compared to the control (5 out of 40 workers; Fisher’s exact test: p < 0.001; Fig. 1). The filtered MP was subsequently ejected as a compact MP pellet (see Fig. A7). We found multiple pellets and pellet fractions on refuse piles and elsewhere within the nest boxes in the MP treated microcolonies and none in the control microcolonies. Six workers from the fine MP treatment had fluorescent particles in their digestive system following to the IBP, but none of the control workers (Fisher´s exact test, p = 0.026; Fig. 1). Therefore, workers from microcolonies exposed to fine MP appeared to filter out most MP using their IBP, as evidenced by the lack of MP in the digestive system posterior to the IBP and the presence of MP in the IBP of every worker. The IBP are often completely filled with MP (Fig. 2), and appear in a pink colour using light microscopy (Fig. 2c) or dark after the tissue clearing (Fig. 2g), and exhibit strong fluorescence (Fig. 2d and 2h). No larvae of both treatments contained any MP, and only one pupa had a single fluorescent particle in its oral cavity (control vs fine MP; Fisher´s exact test: both p = 1; Fig. 1). These results indicate that the IBP is an effective, although not perfect, filtration device, that prevents the transfer of MP from workers to larvae in C. japonicus.

The B. terrestris workers did not discriminate against the MP suspensions and frequently fed on them in addition to the sugar water (personal observation; visible in Fig. A5 c). The detection-frequency of MP particles in the digestive systems of bumblebee workers differed significantly among treatments (Pearson's chi-squared test: p < 0.001). In 36 out of 48 workers from colonies provided with fine MP suspension, we found MP particles in their digestive system (Fig. 3a; Fig. A8 e & f). In contrast, none of the digestive systems of workers (n = 48) fed with coarse MP suspension contained MP particles (Fig. 3a; Fig. A8 g & h). In 38 out of 47 workers from colonies provided with TWP suspension, we found TWP particles in their digestive system (Fig. 3a; Fig. A8 i & j). None of the digestive systems of the workers from control (n = 48) or solvent control (n = 48) colonies contained any MP particles (Fig. 3a; Fig. A8 a – d).

The detection-frequency of MP particles in the digestive systems of bumblebee larvae was significantly different among treatments (Pearson's chi-squared test; p < 0.001). We found MP particles in the digestive systems of 47 out of 48 larvae from colonies provided with fine MP suspension (Fig. 3b; Fig. A9 a-c; Fig. A10 e & f). Only two out of 47 larvae from colonies provided with coarse MP suspension had MP particles in their digestive system (Fig. 3b; Fig. A9 d-f; Fig. A10 g & h). We found TWP in the digestive systems of 44 out of 47 larvae from colonies provided with TWP suspension (Fig. 3b; Fig. A9 g-i; Fig. A10 i & j). None of the larvae from control (n = 48) or solvent control (n = 48) colonies had any MP particles in their digestive system (Fig. 3b; Fig. A10 a – d).

Over all colonies and treatments, we did not find any MP particles in the inner body of any of the bumblebee pupae analysed (control, solvent control, TWP: n = 24; fine/coarse MP: n = 23; Fig. A11).

The infrabuccal pocket (IBP) is a filtering device of some ant species such as C. japonicus, which can filter out particulate matter such as MP from food before it enters the crop (Eisner and Happ 1962; Richter and Economo 2023; Le Hen et al. 2024). Due to its filtering capabilities, it could therefore protect individual workers from potential adverse effects of MP ingestion. However, it is unclear whether MP are completely filtered out or if some particles can pass the IBP and reach the crop and / or midgut. In our study all ant workers fed with fine MP collected MP particles in the IBP, while only 15% of these workers also had MP in their digestive system following the IBP. For example, workers of C. pennsylvanicus were able to completely filter out particles > 150 µm (Eisner and Happ 1962), while the red imported fire ant Solenopsis invicta can filter out almost all particles ≥ 0.88 µm (Glancey et al. 1981). In addition, we found that the workers regurgitated the collected MP as compact MP pellets, cleaning the IBP from MP and restoring its filter capacity. Therefore, the IBP likely decreases the risk of adverse effects of MP taken up with food since most particles are filtered out before reaching the digestive system.

In contrast, bumblebee workers do not have a structure equivalent to the IBP. Accordingly, we frequently found MP particles in their digestive system. Unlike ants (Wang et al. 2019; Le Hen et al. 2024), bumblebees are therefore most likely not able to filter out particulate pollution from feeding suspensions before it enters their digestive system. Adverse effects, such as alterations in the gut microbiome composition and gene expression, have previously been documented in bumblebees after ingestion of particles with health risk potential (diesel exhaust particles, Seidenath et al. 2023). Therefore, bumblebee workers may have a health disadvantage compared to ants due to the lack of the IBP. Furthermore, the effects could extend beyond the worker level, as we found MP particles to also be incorporated into the colony matrix (honey and wax, see Fig. A12 for honey and Fig. A13 for wax), as observed in other bee species (e.g. Apis mellifera; Alma et al. 2023).

In bumblebees, we compared three different MP treatments (fine MP, coarse MP, TWP) and found that MP uptake of workers varied greatly between treatments. While fine MP particles and TWP were found in the bumblebee workers’ digestive system with high frequency, no worker contained coarse MP particles in their digestive system. The size class of particles we used for the fine MP treatment (d₁₀ = 26.86 µm, d₉₀ = 127.1 µm) overlaps extensively with the size of pollen of plants that B. terrestris pollinates (Velthuis and Cobb 1991; Free 1993; Rasmont et al. 2008; Kämper et al. 2016; Raine et al. 2022). This size overlap with pollen may hamper the discrimination of fine MP particles from pollen based solely on size, and with it a discrimination against fine MP polluted food during foraging.

In contrast, we did not find coarse MP in the digestive systems of B. terrestris workers. In preliminary tests on the maximum particle size that workers can ingest, we found that B. terrestris workers were able to ingest MP particles up to a diameter of about 405 µm (Fig. A14). As particles of this size were the largest we tested, it is possible that even larger particles can still be ingested orally by bumblebee workers. Thus, it is unlikely that mechanical or morphological filters prevented the uptake of the coarse MP particles during the experimental period due to their size (with 90% of the particles being smaller than 303.6 µm (d₉₀) and d₁₀ = 95.74 µm). In contrast, TWP with a size range (d₁₀ = 94.41µm, d₉₀ = 214.6 µm) overlapping with that of coarse MP, were surprisingly frequently taken up bumblebee workers, even slightly more frequently than the fine MP. Thus, the size of the particles cannot explain these differences. We suggest that other physico-chemical properties such as the texture and surface structure, or surface charge of the TWP may explain these differences in uptake. While the PS particles are solid and with a relatively smooth surface (Fig. A15, a), the TWP are sponge-like and softer with a rough surface (Fig. A15, b). This could make it difficult to recognise the TWP in the feeding suspension, as they may soak up sugar water. The assumed ability of TWP to soak up sugar water is supported by our observation of initially floating TWP sinking down after incubation in sugar water for several hours (personal observation). As a consequence, TWP may pose a higher risk to bumblebee workers as they are taken up more easily than the pristine MP particles we used. Likewise, aged MP particles with more rugged surfaces, altered surface charge, and / or an ecocorona may also be taken up more readily.

Since the IBP is located anterior to the crop, it may prevent the trophic transfer of MP to brood provided from crop content during trophallaxis. In our study, we could not find any MP in the digestive systems of ant larvae, even though all workers fed with fine MP contained MP in their IBP and in some workers, fine MP had reached their gut. This suggests that ant workers do not (or very rarely) transmit MP to their brood during trophallaxis, which protects ant larvae from direct negative effects of MP particles.

In contrast, we found MP in the digestive systems of bumblebee larvae in all three MP-treatments (fine MP, coarse MP, TWP). The frequency varied significantly between the three treatments, with fine MP particles and TWP being present in over 90% of investigated larval digestive systems and coarse MP particles in only 4%. Since larvae are fed by workers this finding suggests that coarse MP must be ingested by workers before it is regurgitated with liquid food for feeding larvae. This may happen at a very low frequency though. The difference in detection frequency of the three MP types in larvae may be explained by the differences in uptake of the MP types by workers, with coarse MP being rarely ingested and fine MP and TWP being ingested frequently. However, these findings are particularly important in the context of a recent study on metal avoidance during cooperative brood care in bumblebees (Gekière et al. 2025). In this study, bumblebees appear to recognise metal pollution in food. While workers not performing brood care ingest the food, they avoid feeding it to the brood. In contrast, bumblebees in our experiment do not seem to recognise MP contamination and do not avoid passing on MP contaminated food during cooperative brood care, even though they had the opportunity to do so in our experimental set-up (see Fig. A5 & A6). This indicates the particular hazard potential of MP pollution in the environment.

A transfer of MP from workers to larvae during cooperative brood care has also been observed in Apis mellifera by Alma et al. (2023). Together with our results, this suggests that the transfer of particulate pollutants from workers to brood during cooperative brood care could be common among bees due to their presumed inability to filter out most particles from liquid food. The consequences of MP transfer from bumblebee workers to larvae can be versatile. Larvae could experience direct adverse effects of MP such as higher mortality (Al-Jaibachi et al. 2019; Fudlosid 2021; Muhammad et al. 2021; Buteler et al. 2022) or sublethal effects such as changes in growth (Fudlosid 2021), alterations in the expression of genes related to immunity and or detoxification (Muhammad et al. 2021; Wang et al. 2021), or increased susceptibility towards other environmental contaminants (Cho et al. 2020). Furthermore, pollution of food with indigestible particles may result in food dilution and subsequent negative effects on growth or development of larvae, as previously observed for other organisms (Amariei et al. 2022; Bucci et al. 2024). Detrimental effects could be carried over into the pupal and / or adult stage and result in smaller, weaker or otherwise negatively affected workers. Such carry-over effects could also negatively impact colony health and functioning (Choe and Rust 2008; Siviter et al. 2018; Schläppi et al. 2021; Stuligross and Williams 2021). Therefore, the transfer of pollutants from workers to larvae may be a disadvantage that could have far-reaching ecological consequences on these important pollinators.

Since we did not find any MP in ant larvae, it is unsurprising that only a single ant pupa contained MP particles. Apart from the IBP in workers, ants purge their digestive system as last instar larvae, which is visible as the meconium after cocoon formation in those species that form a cocoon (Gotoh et al. 2023). This potentially liberates the pupa from previously ingested MP. A similar mechanism is also known in bumblebees, such as B. terrestris. To prepare for the prepupal stage, the larvae gradually stop feeding but continue defecation until their digestive system is completely emptied (Pereboom 2001).

Remaining MP could cause severe problems during the metamorphosis and negatively affect the future worker’s health. For midges the transfer of MP (2µm in diameter) from larval to adult stages has been shown before (Al-Jaibachi et al. 2018, 2019; Setyorini et al. 2021), with MP particles being retained in the digestive system (Setyorini et al. 2021) or possibly also in Malphigian tubules (Al-Jaibachi et al. 2018, 2019) during metamorphosis. Further, drone fly larvae (Eristalis tenax), that were reared in MP-contaminated water with 5000 polyamide fragments/ml, contained MP in their gut as adults. They also weighed 33% and 60% less than pupae and adults respectively, whose larvae had developed in uncontaminated water (Abdulla et al. 2025). In bumblebees, however, we did not detect MP inside the bodies of pupae across all treatments. Therefore, emptying the digestive system seems to be an effective measure to remove particulate pollutants from the bumblebee body before metamorphosis.

Nevertheless, we want to emphasize, that this mechanism is effective for the particulate part only since there is evidence, that toxic substances can be transferred from MPs into the tissues of organisms (Browne et al. 2013; Avio et al. 2015; Ašmonaitė et al. 2018; Batel et al. 2018; Lu et al. 2018; Qu et al. 2019; Qu et al. 2020). The health effects of such leaching events from MPs into the tissues of terrestrial insects have hardly been researched so far. The negative effects of for instance TWP-leachates on aquatic organisms (Gualtieri et al. 2005; Marwood et al. 2011; Khan et al. 2019), however, indicate a potentially major health risk also for social insects.