Contamination status evaluation and safety aspect exploration for an intensified utilization of edible insects in Kenya

The ContamInsect project deals with the use of insects as a protein-rich source of food and feed in African countries, particularly in Kenya. The focus is on investigating the safety and efficiency of insect use, with particular at-tention being paid to the contamination risks posed by mycotoxins, polychlo-rinated dibenzo-p-dioxins and dibenzofurans (PCDD/F), polychlorinated bi-phenyls (PCB) and polycyclic aromatic hydrocarbons (PAH). Mycotoxins are one of the major concerns affecting global food and feed safety. There is opportunity however for contaminated grains or market waste to re-enter the food chain through bioconversion. Black soldier fly (BSF) larvae - used as a rich protein source in animal feed – are stipulated to efficiently grow on contaminated feed without accumulation of mycotoxins, subsequently allowing the larvae to be used as animal feed with minimal food safety issues. To test this, a study was conducted in which BSF larvae were exposed to mixtures of clean and naturally mycotoxin contaminated maize as well as market waste in defined ratios. Thereupon mycotoxin con-centration in feed and insects was quantified. As wild-caught insects play a significant role in the nutrition of the popula-tion in Kenya representative samples were tested for PAH, PCDD/F and PCB contamination. The objective is to examine whether insects can be used as a safe and sus-tainable source of protein without any health concerns arising from the bio-conversion of contaminated feed. Methods for the safe use of highly myco-toxin-contaminated grain by processing it using black soldier fly larvae are to be developed. Public education and procedural instructions should pro-mote the acceptance and use of edible insects as a sustainable source of protein. The analysis of wild-caught insects should provide an overview of possible contamination and the possibility of accumulation of process contaminants such as polyaromatic hydrocarbons (PAH) and persistent organic pollutants (Pshort forphosphorus

Within this project insect and feed samples were screened for 16 mycotoxins which had been found in amounts above the limit of quantification in feed samples previously screened for 76 mycotoxins. Of the 16 mycotoxins measured via LC-MS/MS the 3 most abundant ones found in all feed ratios and all insect samples exposed to mycotoxin containing feed mixtures were beauvericin (BEA), deoxynivalenol (DON) and zearalenone (ZEN). Most abundant corresponding metabolites and modified forms were deoxynivalenol-3-acetate (DON-3-Ac), deoxynivalenol-15-acetate (DON-15-Ac) and zearalenol-alpha (ZEL α). Hardly any mycotoxin contamination could be observed in clean maize and pure market waste, but contaminated maize mixed with market waste showed significantly higher mycotoxin concentrations than mixtures with clean maize. Thus, market waste might provide a substrate for fungal growth resulting in mycotoxin production, while the contaminated maize might carry various spores or fungi of the genera Beauveria and Fusarium. As key finding for DON, DON-3-Ac, DON-15-Ac and ZEN, BSF larvae showed a significant reduction of toxin content compared to the feed, whereas for the more lipophilic compound BEA an accumulation was observed. Interestingly a mixture of 75 % market waste and 25 % contaminated maize led to accumulation of the metabolite ZEL-α in BSF. This finding might be caused by metabolization of the corresponding mother compound in the insects or by transfer from the feed into the insects. In addition, samples of wild-caught termites, crickets, dung beetles, grasshoppers and black soldier flies, some of them already prepared, were analysed for 17 relevant PCDD/Fs and 12 dioxin-like PCBs (dl-PCBs) as well as for 6 non-dioxin-like PCBs (ndl-PCBs) using high-resolution gas chromatography coupled with high-resolution mass spectrometry (HRGC-HRMS). No maximum levels for PCDD/F and PCBs in insects have been set in the EU. In order to be able to better categorise the results, the determined levels were compared with the maximum levels for fishery products and infant and baby food. However, an actual assessment of the levels cannot be made on this basis. This would require further investigations into the proportion of such insects in the daily diet across different population groups.
Levels of less than 1 pg/g fresh weight were analysed for the sum of WHO-PCDD/F-PCB-TEQ ub. This is well below the maximum level for fishery products. The levels of ndl-PCB in prepared insects are close to or below the analytically determinable concentration (LOQshort forLimit of quantification). Samples that have not been prepared occasionally show higher levels of ndl-PCB (up to 10 ng/g fresh weight). However, these are also far below the maximum levels for fishery products.A maximum level for WHO-PCDD/F-PCB-TEQ ub of 0.2 pg/g fresh weight is set for baby and infant food which was exceeded by some of the samples analysed. This applies in particular to already prepared termites that were purchased at local markets. Levels for WHO-PCDD/F-PCB-TEQ ub of up to 0.3 pg/g fresh weight were detected.For the determination of PAH levels, insects that were prepared by methods that suggest possible contamination with PAH were prioritized. The results of the analyses revealed higher PAH contents for two samples of smoked grasshoppers (sum of 16 EPA PAH: 912 and 1114 µgshort formicrogram/kgshort forkilogram, sum of 4 EFSAshort forEuropean Food Safety Authority PAH: 62 and 113 µgshort formicrogram/kgshort forkilogram), which exceed European maximum levels ​​for other types of smoked foods by two- to three-fold. Fried grasshoppers were less contaminated, showing levels of the 4 EFSAshort forEuropean Food Safety Authority PAH in the single-digit or low double-digit µgshort formicrogram/kgshort forkilogram range. Only one of the analyzed samples of fried termites had quantifiable amounts of the 4 EFSAshort forEuropean Food Safety Authority PAH (5 µgshort formicrogram/kgshort forkilogram). Termites also showed the lowest levels of contamination with the 16 EPA PAH. The significantly lower PAH 4 maximum level for foods for special medical purposes intended for infants and small children (1 µgshort formicrogram/kgshort forkilogram) was exceeded by all examined grasshopper samples. It should be noted that the data collected on PAH levels is not representative and therefore only of limited significance.