Proximate Composition of Thai and Cambodian Ready-to-Eat Insects

Institute of Food Quality and Food Safety, Hannover University of Veterinary Medicine (TiHo), Hannover, Germany Livestock Development for Community Livelihood (LDC), Phnom Penh, Cambodia Faculty of Veterinary Medicine, Royal University of Agriculture (RUA), Phnom Penh, Cambodia Basic Veterinary and Animal Science Department, Faculty of Veterinary Medicine, Mahanakorn University of Technology (MUT), Bangkok 10530, 0ailand Veterinary Science Clinic, Faculty of Veterinary Medicine, Mahanakorn University of Technology (MUT), Bangkok 10530, 0ailand Department of Animal Production Technology and Fisheries, Faculty of Agricultural Technology, King Mongkut’s Institute of Technology Ladkrabang (KMITL), Bangkok 10520, 0ailand


Introduction
Nowadays, the world can be divided into countries with and without a tradition of entomophagy.
ailand and Cambodia clearly belong to the first set of countries, each with a large number of traditionally consumed species. In these countries, insects are typically consumed as a snack after having been spiced and deep-fried. e extent of information regarding the composition varies with the species [1,2], and while data are typically based on unprocessed samples, actual nutritional data of ready-to-eat insects in these countries are, to our knowledge, scarce or even inexistent. erefore, the aim of this contribution is to present proximal chemical analysis data to a selection of Cambodian and ai insects in the way the consumer actually eats them. Table 1 lists the sampled species. Insect samples were purchased on the occasion of the 2020 IFNext (bringing insect farming to the next level) project meeting in different places of Cambodia and ailand, ranging from traditionally prepared insects (Lethocerus indicus, Tarbinskiellus portentosus, Bombyx mori, and Cybister limbatus) as sold in the local market of Skuon (Kampong Cham, Cambodia) to industrially processed and packaged Omphisa fuscidentalis purchased at a department store in Bangkok. Samples of Gymnogryllus vietnamensis were bought at a farm shop in â Lȗang (Lòpburi, ailand), while the Teleogryllus mitratus sample was derived from the work of a PhD student at LDC in Phnom Penh. For the transport to Germany (as the analysis was planned to be carried out there), O. fuscidentalis and T. mitratus samples were milled in Phnom Penh, and all samples were kept as cool as possible using thermal packs. Upon arrival, samples were frozen at −18°C (Liebherr Comfort NoFrost, Liebherr-Hausgeräte GmbH, Ochsenhausen, Germany) until analysis. All samples but T. mitratus (dried only) and O. fuscidentalis were spiced and deep-fried. Skuon samples were spiced with soy sauce, chili, and pandan leaves, while G. vietnamensis was spiced in a Tôm Khà Gai (chicken galangal coconut soup) style, and O. fuscidentalis was prepared with salt, pepper, and soy sauce. Both commercial products (G. vietnamensis and O. fuscidentalis) claimed to be oil-free.

Materials and Methods
Species were identified, labeled, and submitted for proximate analysis (dry matter, ashes, crude protein, crude fat, crude fibre, nitrogen-free extract, Na, and Cl) following German standard procedures [3]. In this way, dry matter was obtained after drying the sample at 104°C, ashes by combustion at 550°C until weight remained constant, crude fat by Soxhlet, crude protein by Kjeldahl, crude fibre after treatment with diluted acids and bases, and the nitrogen-free extract by subtracting the sum of the previous components from the dry matter value. Sodium was determined using atomic absorption spectroscopy and chlorine via colorimetric titration with silver nitrate. Samples' sizes allowed splitting the B. mori, G. vietnamensis, and T. portentosus samples into two individual batches (n � 2), but each sample was analysed in duplicate.
Romanisation for Khmer and ai terms followed the recommendations of the Geographic Department of the Cambodian Ministry of Land Management and Urban Planning (GD (https://www.eki.ee/wgrs/rom1_km.pdf )) and the Royal ai General System of Transcription (RTGS (https://www.orst.go.th/iwfm_splash.asp)), respectively. e latter was, however, expanded by using tone signs and transliterating the letter "อ" as "or" as done typically in ailand. However, original insect names were conserved to keep the information as precise as possible. Table 2 shows the complete results of the analysis, while Table 3 relates minimum and maximum values to different species. With values between approximately 660 and 980 g/ kg, all samples yielded relatively little moisture. Obviously, the more moisture the sample contained, the more pronounced the differences between the complete sample and the sample on a dry matter base were. In this way, ashes ranged approximately between 25 and 50 g/kg, crude protein between 200 and 530, crude fat between 320 and 600, crude fibre between 40 and 140, nitrogen-free extract between 2 and 160, sodium (Na) between 3 and 16, and chlorine (Cl) between 6 and 14 g/kg (Table 1). While L. indicus values ranged between the extreme values, T. mitratus yielded the highest protein content and one of the lowest fat contents. Lowest protein contents were found in B. mori and G. vietnamensis, while O. fuscidentalis yielded most fat. C. limbatus turned out to be of low fat but high fibre. Most minerals were detected in T. portentosus and B. mori. e manufacturer of the ai commercial products included compositional data on their products' labels, using a serving base (20 or 25 g). When transforming these values to a 100% base, the protein content of O. fuscidentalis was 24%, fat 56%, N 2 -free extract 12%, and fibre 4%, and for G. vietnamensis, they were 35, 35, 15, and 5%, respectively. While most values for O. fuscidentalis largely corresponded with the results of the present paper, G. vietnamensis was lower in protein, but higher in fat and fibre. More fibre was also encountered in O. fuscidentalis.

Discussion
Even on the DM base, the chemical composition of the analysed insects was, as expected, heterogeneous. is variability is due to taxon, feeding, instar, trimming, and processing, among others [4]. Comparing results with the literature (Table 4) was unsatisfying since most references were based on nonprocessed insects. In addition, no data for G. vietnamensis could be found. e composition of the African cricket species "Gymnogryllus lucens" was documented by Magara et al. [5], but this species was moved to another genus, being currently termed as Phonarellus lucens in the Orthoptera Species File (https://orthoptera.speciesfile. org/Common/basic/Taxa.aspx?TaxonNameID�1122521, accessed on February 17 th , 2021). As not all authors analysed for the same parameters, some table cells had to remain in blank.
Köhler et al. [6] determined crude protein in several ai species including B. mori, reporting 26.6% in fresh weight when sampling deep-fried silkworms from a street vendor, which is slightly more than registered in the present study. e same is true for the sodium content (1.3%). Yhoung-aree et al. [7] also analysed deep-fried O. fuscidentalis, recording 3.9% ash content, 26.70% protein, 55.80% fat, and 11.50% carbohydrates (DM), which largely corresponds to these results.
Data about specific minerals were even scarcer. Lumsaed [8] reported 0.65% salt in both L. indicus and T. mitratus, and apa [10] reported 609.2 mg in O. fuscidentalis. Table 4 also shows marked intraspecies variances, possibly due to analytical differences. Results for T. portentosus included both whole insects and degutted ones. In this way, differences between the present values and the references were expected.
Since Table 4 contains the reference for unprocessed samples and Table 2 the measured values of processed ones, the influence of processing can be estimated. is, however, can only be hypothesized because, on the one hand, the reference values vary strongly, and, on the other hand, the present contribution only measured processed insects rather than following the process chain from specific batches from raw to deep-fried. Still, it was seen that deep frying reduced moisture and protein, increased fat, and seemingly did not affect ash, fibre, nor carbohydrates. Moisture was reduced because of evaporation during frying, and protein may have abandoned the insect body during frying. In fact, personal observations while deep frying insects showed that, after a certain time, a dark sediment forms at the bottom of the frying pan or wok. Another explanation may be that the insects were cooked before deep frying. is also reduces protein and fat contents in the insects [4]. Unlike protein, fat is replaced during deep frying. For a statistical approach, differences between reference and measured values expressed as percentages were compared among each other using Student's t-test when possible (data not shown). Significant (p < 0.05) differences among species occurred for moisture but generally not for the other parameters. Due to gaps in the reference, not all species-parameter combinations could be evaluated. In this way, moisture was reduced between 58 (C. limbatus) and 92% (T. mitratus). Some 19.78 ± 22.32% protein was lost, while fat increased by 226.59 ± 230.34%.
Few data exist on the compositional changes of southeastern Asian insects during processing. Yhoung-aree [21] gathered protein and fat contents of raw, blanched, and fried L. indicus, T. portentosus, B. mori, O. fuscidentalis, and C. limbatus to a varying degree. Using these data for further calculation, it was seen that blanching reduced the protein content by approximately 30% (20% in B. mori), while frying previously blanched insects resulted in an analytical increase by 15%, probably due to water evaporation. Blanching raw insects reduced the fat content to a variable degree (to approximately 84% of the original amount in B. mori and to 48% in T. portentosus), and frying increased it for obvious reasons by approximately 123 to 171%. is tendency was reflected in the present data as protein contents of the  processed insects were typically below the references in Table 4 and fat ranged high above them.

Conclusions
While some species of edible insects are well documented, many are not, particularly traditional ones and particularly in terms of foodstuffs derived from them. In this way, the present short communication sought to help filling gaps. Apart from natural factors (taxon, feeding, etc.), processing adds another dimension to compositional changes. While the values assessed as such do not reflect the references (based on unprocessed specimens), processing is responsible for protein and fat content changes. e tested species largely consisted of protein and fat (which are both very valuable nutrients), while carbohydrates were low. In contrast, potato crisps and other plant-based, deep-fried snacks typically contain more carbohydrates and fat rather than protein.
is combination is beneficial for cases where nutrients have to be supplied to combat malnutrition as it is for the case where obesity is reduced by ketogenic diets.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.