**Growth Performance, Waste Reduction Efficiency and Nutritional Composition of Black Soldier Fly (***Hermetia illucens***) Larvae and Prepupae Reared on Coconut Endosperm and Soybean Curd Residue with or without Supplementation**

**Nichaphon Pliantiangtam 1, Pipatpong Chundang <sup>2</sup> and Attawit Kovitvadhi 2,\***


**Citation:** Pliantiangtam, N.; Chundang, P.; Kovitvadhi, A. Growth Performance, Waste Reduction Efficiency and Nutritional Composition of Black Soldier Fly (*Hermetia illucens*) Larvae and Prepupae Reared on Coconut Endosperm and Soybean Curd Residue with or without Supplementation. *Insects* **2021**, *12*, 682. https://doi.org/10.3390/ insects12080682

Academic Editors: Man P. Huynh, Kent S. Shelby and Thomas A. Coudron

Received: 7 July 2021 Accepted: 27 July 2021 Published: 29 July 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

**Simple Summary:** Black soldier fly (BSF, *Hermetia illucens*) larvae have a high potential to convert organic waste into high-value products. However, the growth performance, waste reduction efficiency, and chemical composition of BSF larvae are greatly influenced by the rearing substrate. This study focused on investigating the growth performance, waste reduction efficiency, and nutritional composition of BSF larvae reared on different ratios of coconut endosperm (C) and soybean curd residue (S), with or without supplementation, compared to standard diets (Gainesville: G and starter chicken diet: CK). The results showed that BSF larvae fed CK has the highest larval weight, followed by those fed coconut endosperm and soybean curd residue at a ratio of 20:80 (C20S80), and coconut endosperm and soybean curd residue at a ratio of 50:50 (C50S50) without supplementation. The greatest waste reduction efficiency was observed in the G, C50S50, and C20S80 groups without supplementation. The highest crude protein content in larvae was presented in the G and C20S80 groups followed by the CK and C50S50 groups. Therefore, equal proportions of C and S without supplementation is likely the best formulation for growth performance, waste reduction efficiency, and nutritional composition of BSF larvae when compared with standard diets.

**Abstract:** Black soldier fly (BSF, *Hermetia illucens*) larvae are considered as insects with a high potential to convert organic waste into high-value products. The objective of this study was to investigate the growth performance, waste reduction efficiency, and nutritional composition of BSF reared on different ratios of coconut endosperm (C) and soybean curd residue (S), with or without supplementation, compared to standard diets (Gainesville: G and starter chicken diet: CK). Sevenday-old larvae were randomly divided into eight experimental groups (G, CK, and three different ratios of C and S with or without supplementation) with three replicates with an equal weight of larvae. The supplement contained calcium, phosphorus, amino acids, and a mineral–vitamin premix which was formulated to correlate with CK. Each replicate was terminated, measured, and evaluated when 40% of larvae had reached prepupal stage. The highest larval weight gain was presented in BSF fed CK, followed by those fed coconut endosperm and soybean curd residue at a ratio of 20:80 (C20S80), and coconut endosperm and soybean curd residue at a ratio of 50:50 (C50S50) without supplementation (numbers after C and S represent their percentage in the formulation; *p* < 0.001). Harvesting was delayed in the BSF fed C80S20 with and without supplementation (*p* < 0.001). The number of total larvae and prepupae was not significantly different between groups (*p* > 0.05). The greatest waste reduction efficiency was observed in the G, C50S50, and C20S80 groups without supplementation (*p* < 0.001). All groups with supplementation had a higher proportion of ash in both larvae and prepupae compared to non-supplemented groups (*p* < 0.001), but lower growth performance. The highest percentage of crude protein in larvae was presented in the Gainesville and C20S80 groups followed by the CK and C50S50 groups (*p* < 0.001). Equal proportions of C and S without supplementation are suggested as a rearing substrate. However, growth performance was

lower than for CK; therefore, further studies could investigate cost-efficient techniques to promote this parameter.

**Keywords:** *Hermetia illucens*; organic waste management; coconut endosperm; soybean curd residue

#### **1. Introduction**

The world population has increased sharply in recent decades and could reach 9.7 billion in 2050 [1]. As a consequence, there is a higher demand for food. One-point-three billion tons of food waste is estimated to be generated per year following the sharp increase of consumption [2]. This large amount of organic waste is mainly sent to the landfill [3]. The gas released from this landfill contributes to the greenhouse effect and global warming [4,5]. It is clear from these problems that it is economically sound to study and conduct appropriate management of organic waste.

Several insects have the potential to decompose organic waste and convert it into biomass [6]. One of the most interesting insects to use as a professional decomposer is the black soldier fly (BSF), *Hermetia illucens* (Diptera: Stratiomyidae) [5–7]. BSF is considered a large member of the order Diptera at around 15–20 mm long [8]; it lives in the tropical zone [9], is not a disease vector, and is not harmful to humans or animals [10]. BSF larvae (feeding stage) can efficiently decompose several types of organic waste including poultry manure [11,12], cow manure [12,13], swine manure [12], human feces [13], pig's liver waste [14], fish industrial waste [14], poultry industrial waste [13], restaurant waste [13–15], vegetable waste [13,15], fruit waste [16], vegetable and fruit waste [14,16,17], pineapple and jackfruit peel [18], wheat bran [13,19], maize straw [19], and beer and wine by-products [16]. In addition, BSF larvae can convert this organic waste into high-value products: protein and lipid sources for the animal feed industry [20], biodiesel [21], and antimicrobial peptides [22]. Therefore, BSF rearing could be a solution to managing and upcycling organic waste in an environmentally friendly and economically sustainable way. However, the growth performance, waste reduction efficiency, and chemical composition of BSF are greatly influenced by the rearing substrate [4,5,15]. Based on this knowledge, an appropriate rearing substrate could be used to achieve the highest decomposition efficiency and good-quality end products.

Coconut endosperm (C) and soybean curd residue (S) were used as a rearing substrate in a recent study by Lim et al. [21]; the chemical composition of rearing substrates (crude protein 8.18–20.2% DM (dry matter) and lipid 31.2–31.5% DM) that they reported was quite different from that in our study (crude protein 4.35–11.2% DM and lipid 4.69–5.61% DM), even though the same industrial by-products were used. C and S are easy to obtain in the local market or at industrial scale, as they are considered common by-products. Therefore, C and S were selected for use in this study. In most studies, good chemical composition of BSF was achieved by using chicken diet as a substrate compared to organic waste, but it is not economically sound in reality [15,23] Therefore, supplementation of organic waste with calcium, phosphorus, essential amino acids, and vitamin–mineral premix, to make it similar to chicken diet, could support the performance and quality of BSF. Based on this hypothesis, the objective of this study was to investigate the growth performance, nutritional composition, and waste reduction efficiency of BSF by rearing flies on substrates containing different ratios of C and S, with or without supplementation, compared to the standard diets (Gainesville and starter broiler chicken diet).

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

#### *2.1. Insects, Rearing Substrates, and Chemical Analysis*

Seven-day-old larvae were randomly collected from a colony of BSF larvae (Orgafeed Co., Ltd., Bangkok, Thailand) which were reared on starter broiler chicken diets (Table 1). All larvae were randomly assigned into eight experimental groups with three replicates per

group (2.82 g or approximately 200 larvae per replicate). The larvae in each experimental group were fed different diets: 1. Gainesville diet (G; Scala et al. [5]); 2. Starter broiler chicken diets (CK); 3. Coconut endosperm and soybean curd residue at a ratio of 80:20 (C80S20); 4. Coconut endosperm and soybean curd residue at a ratio of 50:50 (C50S50); 5. Coconut endosperm and soybean curd residue at a ratio of 20:80 (C20S80); 6. Coconut endosperm and soybean curd residue at a ratio of 80:20 with supplementation (C80S20s); 7. Coconut endosperm and soybean curd residue at a ratio of 50:50 with supplementation (C50S50s); 8. Coconut endosperm and soybean curd residue at a ratio of 20:80 with supplementation (C20S80s). The G diet was accepted as the general experimental diet for insects in Diptera and was used as control diet in several research studies [4,5,8,12]. Therefore, the G diet was used as the control diet to compare between groups in this and other studies. The ingredients and chemical composition of diets and supplements were evaluated based on proximate analysis (AOAC 2006) and are shown in Table 1. The supplement was formulated correlating to the macro-minerals, vitamins, micro-minerals, and amino acids in the CK diet. The C and S were obtained as industrial by-products from Dutch Mill Co., Ltd. (Bangkok, Thailand). The C was treated by anaerobic fermentation for 4 weeks prior to usage [21].

**Table 1.** Ingredients and chemical composition of experimental diets.


<sup>1</sup> Gainesville diet contains wheat bran, alfalfa meal, and corn meal at 50%, 30%, and 20%, respectively. <sup>2</sup> Chicken diet (Starter) contains corn meal, soybean meal, palm oil, monodicalcium phosphate, limestone, salt, vitamin–mineral premix (Feed specialties Co., Ltd.; Pathumthani, Thailand), DL-methionine, sodium bicarbonate, and choline chloride at 50.8%, 39.3%, 6.06%, 1.38%, 1.38%, 0.24%, 0.36%, 0.25%, 0.20%, and 0.07%, respectively. <sup>3</sup> Five grams of supplement contains monodicalcium phosphate, lime stone, DL-methionine, L-lysine, L-threonine, lard, and vitamin–mineral premix (Feed specialties Co., Ltd.; Pathumthani, Thailand) at 2.2 g, 1 g, 0.3 g, 0.14 g, 0.06 g, 1 g, and 0.3 g, respectively. Vitamin–mineral premix (Feed specialties Co., Ltd.; Pathumthani, Thailand) were supplied per kilogram of diets at 2,500,000 IU of vitamin A; 1,000,000 IU of vitamin D3; 7000 IU of vitamin E; 700 mg of vitamin K; 400 mg of vitamin B1; 800 mg of vitamin B2; 400 mg of vitamin B6; 4 mg of vitamin B12; 30 mg of biotin; 3111 mg of Ca pantothenate acid; 100 mg of folic acid; 15,000 mg of vitamin C; 5600 mg of vitamin B3, 10,500 mg of Zn, 10,920 mg of Fe; 9960 mg of Mn; 3850 mg of Cu; 137 mg of I; 70 mg of Se.

#### *2.2. Rearing, Data Collection, Chemical Analysis, and Calculation*

The larvae were placed in a plastic container (15 cm × 24.5 cm × 6.5 cm) on the rearing substrates which were adjusted to obtain an equal humidity of 70% by analyzing the moisture in each substrates and adding water to reach the equal humidity based on calculation before providing into the container. Controlled temperature (28 ± 2 ◦C) and a dark room were used in this study. Each rearing container was checked twice daily at 09:30 and 16:30. Diet was added into the rearing container to achieve sufficient diet during the experiment. Each replicate was terminated when 40% of the larvae had developed into the prepupal stage [14,19]. The amount of substrate added, amount of substrate remaining, larval weight, prepupal weight, number of larvae, and number of prepupae were measured. Moreover, the pH of rearing substrates was measured at the beginning and end of the experiment by mixing rearing substrates with distilled water at 1:10 *w*/*v* [21]. The remaining substrate, larvae and prepupae were frozen and kept at −20 ◦C for further analysis. The larvae and prepupae were dried at 60 ◦C for 48 h and ground into a powder by passing through a 1-mm sieve to identify the DM, crude protein (CP), ash, and ether extract (EE), whereas the remaining rearing substrate was evaluated by DM (AOAC 2006). Substrate reduction (%SR), waste reduction index (WRI), and efficiency of conversion of digested food (ECD) were calculated following Meneguz et al. [16] and represented the formulation as below. Larval weight gain was calculated by dividing the increment in total larval weight between 7 and 14 days by seven. This study was carried out following the standard guidelines approved by the Institutional Animal Care and Use Committee of Kasetsart University, Bangkok, Thailand (ACKU63-VET-004).

$$\% \text{SR} = \frac{\text{Distributed substrate} (\text{g}) - \text{Residual substrate} (\text{g})}{\text{Distributed substrate} (\text{g})} \times 100\tag{1}$$

$$\text{WRI} = \frac{\left\{ \frac{\text{Distributed substrate} (\text{g}) - \text{Residual substrate} (\text{g})}{\text{Distributed substance} (\text{g})} \times 100 \right\}}{\text{Days of trial} (\text{day})} \tag{2}$$

$$\text{ECD} = \frac{\text{Lurability and prepump weight} (\text{g})}{\text{Distributed substrate} (\text{g}) - \text{Residual substrate} (\text{g})} \tag{3}$$

#### *2.3. Statistical Analysis*

This experiment was performed under a completely randomized design. One-way analysis of variance (ANOVA) was performed to evaluate the differences in all measured, analyzed, and calculated data between experimental groups (fixed factors) by using Duncan's multiple range test as post-hoc analysis. The normal distribution and homogeneity of variance were confirmed by the Shapiro–Wilk test and Levene's test, respectively. Statistically significant difference was accepted at *p* < 0.05. All statistical analyses in the study were investigated by using the R statistics program: RStudio v1.4.1103 with the Rcmdr package (R Development Core Team 2008).

#### **3. Results**

The growth performance, chemical composition, waste reduction efficiency, and rearing substrate pH of BSF reared on mixed industrial by-products compared with G and CK are presented in Table 2. The highest larval weight at 14 days and larval growth rate were observed in the CK group followed by the C20S80 and C50S50 groups; the lowest performance was found in the C80S20s group (*p* < 0.001). The latest harvesting date was found for the larvae fed C and S at a ratio of 80:20 with and without supplementation; the harvesting period was around 10–11 days for other groups (*p* < 0.001). The C80S20s group had the lowest final total larval weight (*p* < 0.05). The lowest weight of each larva was presented in the C80S20s group (*p* < 0.001). There was no statistically significant difference in the total number of larvae and prepupae between groups (*p* = 0.08); however, the number of prepupae in the CK and C20S80 groups was higher than in other groups (*p* < 0.001). All groups fed C/S without supplementation (C80S20, C50S50 and C20S80) as well as the C80S20s and G groups had a higher %SR than the CK and C50S50s groups; the lowest was presented in C20S80s (*p* < 0.001). In addition, the C50S50, C20S80, and G groups had a higher WRI than the CK, C80S20, C80S20s, and C50S50 groups; the lowest was presented in C20S80s (*p* < 0.001). The greatest ECD was found in C20S80s followed by C50S50, C80S20, C20S80, C50S50s, CK, G, and C80S20s, respectively. There was a large variation in substrate pH between the beginning of the experiment and the harvesting date: 4.05–6.42 and 4.78–7.17, respectively. The lowest substrate pH at the beginning was presented in C80S20, followed by C50S50, C80S20s, C20S80, C50S50s, C80S20s, G and CK groups, respectively. Substrate pH at the end of the experiment was higher than at the

beginning. The most basic substrate condition at the end of the experiment was observed in the G group followed by the CK group compared to the others (*p* < 0.001).

**Table 2.** Growth performances, waste reduction efficiency, and rearing substrate pH of black soldier fly reared on mixed industrial by-products comparing with Gainesville and chicken diet.


<sup>1</sup> The differences on superscripts in the same row represent the statistical significant difference at *p* < 0.05. <sup>2</sup> Each diet contains five grams of supplement comprising monodicalcium phosphate, lime stone, DL-methionine, L-lysine, L-threonine, lard, and vitamin–mineral premix (Feed specialties Co., Ltd.; Pathumthani, Thailand) at 2.2 g, 1 g, 0.3 g, 0.14 g, 0.06 g, 1 g, and 0.3 g, respectively. <sup>3</sup> Total larvae and/or prepupae were used to calculate these parameters in each study group. <sup>4</sup> Larval weight gain in fresh matter was calculated between 7- to 14 day-old larvae. Vitamin–mineral premix (Feed specialties Co., Ltd.; Pathumthani, Thailand) were supplied per kilogram of diets at 2,500,000 IU of vitamin A; 1,000,000 IU of vitamin D3; 7000 IU of vitamin E; 700 mg of vitamin K; 400 mg of vitamin B1; 800 mg of vitamin B2; 400 mg of vitamin B6; 4 mg of vitamin B12; 30 mg of biotin; 3111 mg of Ca pantothenate acid; 100 mg of folic acid; 15,000 mg of vitamin C; 5600 mg of vitamin B3, 10,500 mg of Zn, 10,920 mg of Fe; 9960 mg of Mn; 3850 mg of Cu; 137 mg of I; 70 mg of Se.

> The chemical composition of BSF larvae and prepupae reared on mixed industrial byproducts compared with G and CK is presented in Figure 1 and Supplementary Table S1. Based on the percentages of nutrients, ash content was lower in all groups fed C and S without supplementation, in both larvae and prepupae, compared to other groups (*p* < 0.001). The highest ash percentage was present in the groups fed C and S with supplementation; the highest was observed in the C20S80s group (*p* < 0.001), similar to that in CK and G groups. A high CP content in larvae and prepupae was observed in the G and C20S80 groups followed by the C50S50 and CK groups, whereas in other groups, it was lower. The lowest fat proportion in larvae and prepupae was present in the G group (*p* < 0.001). In contrast, the highest fat composition was present in the C80S20 group in both larvae and prepupae (*p* < 0.001).

**Figure 1.** Dry matter ((**A**), %FM), ash ((**B**), %DM), crude protein ((**C**), %DM), and ether extract ((**D**), %DM) of black soldier fly larvae and/or prepupae reared on mixed industrial by-products (C: Coconut endosperm; S: soybean curd residue; number after the abbreviation represents the ratio of mixed industrial by-products) with or without supplementation (s) comparing with Gainesville and chicken diet. The statistical significant difference (*p* < 0.05) is represented by the difference of small capital letters inside bar between the experimental groups of larvae (black color) or prepupae (white color).

#### **4. Discussion**

The chemical composition of rearing substrates is considered a major factor influencing the variation in growth performance, waste reduction efficiency, and nutritional composition of BSF larvae and prepupae [14,16,21,24,25]. Lim et al. [21] determined that the total final weight of larvae and growth rate were positively correlated with the amount of CP in the rearing substrate; the highest growth performance was present in larvae fed a mixture of C and S (60:40) containing 12.44% CP. The results of this study agree with those of Lim et al. [21], because the CK diet containing a high CP content (21.8%) provided a higher larval weight at 14 days and greater larval weight gain than in groups fed C and S (4.35–11.2%). Interestingly, Lim et al. [21] found that an excess CP level in the rearing substrate led to negative outcomes for these parameters, which was also represented in this study, because excess protein intake results in energy loss from metabolism involving the excretion of toxic nitrogenous waste [25]. Interestingly, the different chemical composition of the same raw materials between this study and Lim et al.'s [21] was present, but the consequences on growth performances and larval weight gain were similar. Therefore, the cause of this consequence could be influenced by several factors which could be interesting to study. However, CP content could not be the single factor which influences growth parameters. Poor BSF growth performance is found when using rearing substrates with a low caloric density, i.e., low fat and/or carbohydrate, because these nutrients serve as energy sources [23]. Generally, an appropriate ratio between CP and metabolizable energy must be formulated to obtain the highest performance in livestock animals [4]. Therefore, appropriate chemical composition of the rearing substrate could be another consideration point for BSF as it is in livestock animals. A 1:1 ratio of CP to carbohydrate for BSF cultivation has been reported to achieve the fastest development [4]. In our study, the CP:carbohydrate ratio of 0.36 in the CK diet provided the highest performance. On the one hand, C50S50 and C20S80 with CP:carbohydrate ratios of 0.23 and 0.40, respectively, presented a higher larval growth rate compared to other groups fed industrial by-products. Poor performance was found in this study when the CP:carbohydrate ratio was 0.14. However, it still cannot be concluded that only this ratio is the primary factor influencing growth rate. In our

aspect, nutrient quality of amino acids and digestibility could be further studied in depth as another consideration point.

There was no significant difference in the total number of larvae and prepupae between groups in this study. We can assume that the mortality rate was not affected by experimental diets. Most studies report that the mortality rate does not change between rearing diets [21], because BSF can survive on poor nutrient diets and large environmental condition [26]. However, the duration from larvae to prepupae is prolonged when larvae are reared on diets lacking certain nutrients, mainly protein, which prolongs the cultivation period [24,25]. The feeding period is prolonged until the nutrients inside the larvae meet the requirements for development and metamorphosis at which stage they can no longer consume feed [24]. The latest harvesting date (22–23 days old) was found in the larvae fed C and S at a ratio of 80:20, which is a low protein proportion compared to that fed to other groups (which were harvested at 17 days old). In another study, the shortest rearing period (19 days old) was presented by larvae fed a mixture of C and S at a ratio of 60:40, containing the highest CP compared to other study groups [21]. In this study, the larvae being fed the CK diet for 7 days before consuming the organic waste could be the cause of an earlier harvesting date compared to the study of Lim et al. [21], in which larvae were reared on organic waste throughout the experiment [13,20]. Therefore, an appropriate starter diet could be used before rearing on organic waste to reduce rearing duration.

The pH condition of substrates did not influence the final weight [27]. However, Ma et al. [28] demonstrated that the initial substrate pH influences the final weight, the best growth performance being observed at pH 6. In the same way, Lim et al. [21] found the highest total weight and growth rate when rearing BSF at pH 5.82. In addition, it has been suggested that the rearing substrate pH should be higher than 6 to achieve a good productive performance [9]. In our study, the substrate with a pH of 6.42 (CK diet) resulted in the significantly heaviest larvae weight at 14 days, whereas a lower larvae weight was found in other groups fed industrial by-products with an initial substrate pH of 4.05–5.36. Therefore, the difference in initial substrates pH could be another cause of the diverse outcomes in this study.

The rearing substrate has a direct impact on the nutritional composition of BSF larvae and prepupae [15,16,25]. The chemical composition of the mixed organic diet and that of the BSF larvae and prepupae in this study demonstrate that the CP content in the rearing diet influences BSF larval and prepupal protein content. A high CP content in the diet produced a high protein content in larvae, a result similar to that of other studies [16]. Nitrogen-free extract (NFE) in the diet is positively correlated to the fat content in BSF larvae [25]. In this study, the increment of NFE in the mixed organic diet promoted a higher fat content in BSF larvae than in another study [15]. In addition, larvae reared on a diet high in soluble carbohydrates (618 g/kg) had the highest fat content (386 g/kg DM), whereas those reared on a diet low in soluble carbohydrates (7 g/kg) had the lowest fat content (218 g/kg DM), because insect larvae can convert excess carbohydrates to fat and store it in their body mass [16].

A great increase of ash composition (around three times) was observed in this study in both larvae and prepupae fed supplemented industrial by-products compared to the non-supplemented group. Increasing the calcium level in the substrate correlates with a higher calcium composition in house cricket (*Acheta domestica*) and yellow mealworm [29]. The supplement mainly contained calcium and phosphorus, which could be the cause of this consequence. Therefore, increasing the mineral content in the rearing substrate promoted an increase of these minerals in BSF larvae and prepupae as it did in house cricket and yellow mealworm [29]. From current knowledge from our and other studies, CK diet promotes good growth performance for BSF [15]. Therefore, the supplement was formulated to supply calcium, phosphorus, amino acids, and vitamin–mineral premix to industrial by-product diets, reflecting the CK diet which can provide benefits. However, the nutrient profiles of CK and industrial by-product diets are quite different. Our results showed that the ash content in BSF larvae and prepupae fed with supplements was similar

to that in those fed CK, but there was a deterioration in growth performance compared to CK. In addition, major nutrients such as CP and carbohydrate could be more important factors influencing the growth performance than the supplement. Therefore, increasing the mineral content in rearing substrates which do not contain as appropriate a nutritive value as CK is not indicated for BSF, because a deterioration in growth performance can occur as in this study.

#### **5. Conclusions**

In this study, our results for growth performance, waste reduction efficiency, and the nutritional composition of BSF larvae reared on C50S50 were very similar to those for BSF larvae reared on standard diets (G and CK). However, the growth performance remained lower than in BSF fed CK. Supplementation is not necessary to improve growth performance, waste reduction efficiency, and nutritional composition when feeding larvae industrial by-product diets. Further studies could investigate the clear nutrient requirements of BSF, correlating them with their performance and quality. A technique for improving the growth performance of BSF, to reach a level similar to or better than that of larvae fed CK, by using industrial by-products as rearing substrates, could also be studied.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/10 .3390/insects12080682/s1, Table S1: Chemical composition of black soldier fly larvae and prepupae reared on mixed industrial by-products comparing with Gainesville and chicken diet.

**Author Contributions:** Conceptualization, A.K.; methodology, N.P. and A.K.; formal analysis, N.P. and A.K.; investigation, N.P., P.C. and A.K.; data curation, N.P. and A.K.; writing—original draft preparation, N.P. and A.K.; writing—review and editing, N.P., P.C. and A.K.; project administration, A.K.; funding acquisition, A.K. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by Thailand Science Research and Innovation and Office of National Higher Education Science and Research and Innovation Policy Council (Thailand). This work was partially supported by the Faculty of Veterinary Medicine, Kasetsart University, Bangkok, Thailand.

**Institutional Review Board Statement:** Not applicable.

**Acknowledgments:** The authors would like to humbly thank Orgafeed Co. Ltd. (Bangkok, Thailand) and Dutch Mill Co., Ltd. (Bangkok, Thailand) for providing insects and rearing substrates. The authors would like to thank Sathita Areerat and Penpicha Kongsup for technical assistance on animal husbandry and sample collection.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


## *Article* **Effects of Different Nitrogen Sources and Ratios to Carbon on Larval Development and Bioconversion Efficiency in Food Waste Treatment by Black Soldier Fly Larvae (***Hermetia illucens***)**

**Yan Lu 1,2,†, Shouyu Zhang 1,2,†, Shibo Sun 3, Minghuo Wu 1, Yongming Bao 1, Huiyan Tong 1, Miaomiao Ren 1, Ning Jin 1, Jianqiang Xu 2,3, Hao Zhou 1,4 and Weiping Xu 1,2,4,\***


**Simple Summary:** Black soldier fly larvae (BSFL) have received global research interest and industrial application due to their high performance on the organic waste treatment. However, the substrate C/N property, which may affect larvae development and the waste bioconversion process greatly, is significantly less studied. The current study focused on the food waste treatment by BSFL, compared the nitrogen supplying effects of 9 nitrogen species (i.e., NH4Cl, NaNO3, urea, uric acid, Gly, L-Glu, L-Glu:L-Asp (1:1, *w*/*w*), soybean flour, and fish meal), and further examined the C/N effects on the larval development and bioconversion process. We found that NH4Cl and NaNO3 led to poor larval growth and survival, while 7 organic nitrogen species exerted no harm to the larvae. Urea was further chosen to adjust the C/Ns. Results showed that lowering the C/N from the initial 21:1 to 18:1–14:1 improved the waste reduction and larvae production performance, and C/N of 18:1–16:1 was further beneficial for the larval protein and lipid bioconversion, whereas C/N of 12:1–10:1 resulted in a significant performance decline. Therefore, the C/N range of 18:1–16:1 is likely the optimal condition for food waste treatment by BSFL and adjusting food waste C/N with urea could be a practical method for the performance improvement.

**Abstract:** Biowaste treatment by black soldier fly larvae (BSFL, *Hermetia illucens*) has received global research interest and growing industrial application. Larvae farming conditions, such as temperature, pH, and moisture, have been critically examined. However, the substrate carbon to nitrogen ratio (C/N), one of the key parameters that may affect larval survival and bioconversion efficiency, is significantly less studied. The current study aimed to compare the nitrogen supplying effects of 9 nitrogen species (i.e., NH4Cl, NaNO3, urea, uric acid, Gly, L-Glu, L-Glu:L-Asp (1:1, *w*/*w*), soybean flour, and fish meal) during food waste larval treatment, and further examine the C/N effects on the larval development and bioconversion process, using the C/N adjustment with urea from the initial 21:1 to 18:1, 16:1, 14:1, 12:1, and 10:1, respectively. The food wastes were supplied with the same amount of nitrogen element (1 g N/100 g dry wt) in the nitrogen source trial and different amount of urea in the C/N adjustment trial following larvae treatment. The results showed that NH4Cl and NaNO3 caused significant harmful impacts on the larval survival and bioconversion process, while the 7 organic nitrogen species resulted in no significant negative effect. Further adjustment of C/N with urea showed that the C/N range between 18:1 and 14:1 was optimal for a high waste reduction performance (73.5–84.8%, *p* < 0.001) and a high larvae yield (25.3–26.6%, *p* = 0.015), while the C/N

**Citation:** Lu, Y.; Zhang, S.; Sun, S.; Wu, M.; Bao, Y.; Tong, H.; Ren, M.; Jin, N.; Xu, J.; Zhou, H.; et al. Effects of Different Nitrogen Sources and Ratios to Carbon on Larval Development and Bioconversion Efficiency in Food Waste Treatment by Black Soldier Fly Larvae (*Hermetia illucens*). *Insects* **2021**, *12*, 507. https://doi.org/10.3390/ insects12060507

Academic Editors: Man P. Huynh, Kent S. Shelby and Thomas A. Coudron

Received: 18 April 2021 Accepted: 11 May 2021 Published: 31 May 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

range of 18:1 to 16:1 was further optimal for an efficient larval protein yield (10.1–11.1%, *p* = 0.003) and lipid yield (7.6–8.1%, *p* = 0.002). The adjustment of C/N influenced the activity of antioxidant enzymes, such as superoxide dismutase (SOD, *p* = 0.015), whereas exerted no obvious impact on the larval amino acid composition. Altogether, organic nitrogen is more suitable than NH4Cl and NaNO3 as the nitrogen amendment during larval food waste treatment, addition of small amounts of urea, targeting C/N of 18:1–14:1, would improve the waste reduction performance, and application of C/N at 18:1–16:1 would facilitate the larval protein and lipid bioconversion process.

**Keywords:** nitrogen source; carbon to nitrogen ratio; food waste; urea; black soldier fly larvae; *Hermetia illucens*

#### **1. Introduction**

Larvae of black soldier fly (BSFL), Diptera:Stratiomyidae, *Hermetia illucens* (Linnaeus), are capable of converting various biowaste into protein-rich insect biomass and nitrogenrich organic fertilizer [1–3] and have received worldwide research interest and fast-growing industrial application [4,5]. The application of BSFL for waste processing has expanded from tropical and temperate countries [4,5] to Russia [6], Canada [7], and Near East Turkey [8] in recent years. Amongst all the parameters that may impact the larval development and biowaste conversion process, ones such as temperature, moisture content, and pH are critically analyzed [9–11]. In contrast, the substrate C/N property, which may play a crucial role in the larval development and bioconversion process, is largely less studied. Several studies have examined the C/N effects indirectly, for instance, Bessigamukama et al. [12], Ewald et al. [13], and Lopes et al. [14] have added different amounts of biochar [12], fish [13], or mussel [14] to the grain or bread substrate for BSFL treatment, and Pang et al. [15] has studied C/N effects on the greenhouse gas emission; however, the C/N effects on larval development are still not clear and the optimal C/N range for a high bioconversion performance has not been achieved, since the substrate total C/N ratios were not determined or the bioconversion efficiencies of BSFL treatment were not examined in those studies.

The BSFL are able to degrade a wide range of organic waste, such as animal manure, food waste, abattoir waste, and aquaculture waste [16–19]. These wastes have different C/N properties, and BSFL are found to be largely more adapted to the C/N range of <20:1 and less suitable to the C/N range of >20:1. For instance, poultry, swine, and human feces, typically having C/N of 9:1 to 15:1, were found to be well-degraded by BSFL, while cow manure, having C/N of 20:1 to 30:1, was less decomposed by BSFL [20,21]. Poultry feed (C/N 18:1), food waste (C/N 14:1), and abattoir waste (C/N 6:1) showed acceptable and comparable decomposition, while fruit and vegetable waste (C/N 24:1) was found to be less processed by the BSFL [18]. In contrast to the animal manure waste, the food waste could be widely different on the C/N properties due to the waste composition. The Chinese diet habit results in a great proportion of carbohydrate (rice, noodles, and steamed buns) left in the food waste, similar to those food wastes containing high amounts of starches [22], which lead the C/N to be occasionally higher than 20:1. This kind of food waste needs to be recycled by the BSFL, which results in an interesting substrate with high C/N properties and arouses practical questions about what nitrogen source should be used for the nitrogen supplement and what C/N range is suitable for a high performance of BSFL treatment.

Both inorganic and organic N species could be examined for the nitrogen supplemental effects on the food waste treatment. Previously, biowaste leachate has been reported to be treated by BSFL [23,24]. Since NH4-N and NO3-N are typical nitrogen species in the leachate, the NH4Cl and NaNO3 could be tested for the nitrogen supplemental performance. Poultry and swine feces are reported to be effectively degraded by BSFL [20,21]. Since urea and uric acid are typical compounds in the feces, these two chemicals could be tested for the nitrogen supplemental efficiency. BSFL protein is found to be rich with L-Glu and

L-Asp amino acids [25]. The L-Glu, L-Asp, as well the simplest amino acid of Gly could be tested for the N supplemental efficiency. Soybean flour and fish meal could be used as positive controls for the organic nitrogen species. As the BSFL are adapted to the poultry manure of C/N 9:1 and less adapted to the cow manure of 20:1 [21,26], the food waste C/N ratio could be adjusted from 20:1 to 10:1 in order to identify the optimal C/N range for a high larval conversion performance.

Therefore, the present study aimed to compare the nitrogen supplying effects of 9 nitrogen species (i.e., NH4Cl, NaNO3, urea, uric acid, Gly, L-Glu, L-Glu:L-Asp (1:1, *w*/*w*), soybean flour, and fish meal) and further examine the C/N effects on the larval development and bioconversion process for the food waste treatment. Larval enzyme activity and amino acid composition were partially analyzed in order to study the potential physiological effects of nitrogen supplement.

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

#### *2.1. Larvae, Food Waste, and Nitrogen Source Preparation*

Black soldier fly eggs were purchased from a BSF farm (Baiaotai farm, Anyou Biotechnology Group Co., Ltd., Guangxi, China). Upon arrival, eggs were hatched for six days in a substrate containing 60% soybean meal, 30% corn powder, and 10% wheat bran in a 65% moisture content environment at 25 ◦C. The 6-day old larvae (average weight 0.0027 g) were removed from the hatching substrate through sieving (1 mm mesh) and 45 batches of approximately 800 larvae were weighed.

Food wastes (FW) were the cooked food leftovers (rice, noodles, vegetables, meats, eggs, etc.) that were collected from the university canteen (Dalian University of Technology, Panjin Campus, Panjin, China). After collection, food wastes were homogenized with a kitchen blender, tested for moisture content in duplicate (oven drying at 105 ◦C until constant weight [27]), and stored at 4 ◦C or −20 ◦C prior to further usage. In Trial 1, 9 nitrogen sources were used, including NH4Cl, NaNO3, urea, uric acid, Gly, L-Glu, L-Asp, soybean flour, and fish meal, while only urea was used in Trial 2. Within the 9 nitrogen sources, 7 pure compounds with purity grade >99% (NH4Cl, NaNO3, urea, uric acid, Gly, L-Glu, and L-Asp) were purchased from Aladdin (Shanghai Aladdin Biochemical Technology Co., Ltd., Shanghai, China), and 2 nitrogen sources (soybean flour and fish meal) were purchased from local stores. Carbon and nitrogen contents of the 7 pure compounds were calculated based on their molecular weight, while the C and N properties of the soybean flour, fish meal, as well as the food waste were determined in duplication using the Vario EL cube elemental analyzer (Elementar Analysensysteme GmbH, Hanau, Germany) with the freeze-dried subsamples (Table 1).

#### *2.2. Experimental Design*

In Trial 1, different nitrogen sources were supplied in 1 g N element per 100 g (dry wt) food waste, i.e., changing N% of food waste from 2.26% to 3.26%. Since the 9 nitrogen sources contained different N and C contents, the actual weights of each nitrogen source used are reported in the Table 1, as well as the C/N values after the nitrogen supplement. In Trial 2, urea was added to the food waste, aiming to adjust food waste C/N from 21:1 (blank control) to 18:1, 16:1, 14:1, 12:1, and 10:1, respectively, while the actual weights of urea used were listed in the Table 1.

All the experiments in Trials 1 and 2 were performed in triplicate, with food waste without nitrogen amendment served as the blank control. Larvae were reared in 4.6 L plastic boxes (240 × 120 × 160 mm) individually. Ten 6 mm diameter holes were made on the box lid in order to enhance passive aeration. In each box, 300 g (wet weight, 70% moisture content) of food waste was added, 150 g food waste was added on Day 0, and another 150 g food waste was added on Day 6. The nitrogen sources were added associated with the food waste according to the amounts in Table 1 on Days 0 and 6, respectively. The 800 weighed larvae (6-days-old) were added into each box on Day 0 following the addition of food waste and nitrogen sources. The boxes were kept at 26–32 ◦C, and the substrates

were mixed manually twice per day. After 12 days, larvae in each box were separated from the frass manually, rinsed with tap water, and dried on paper towels. The total number of larvae was counted. The total wet weight of larvae and frass were recorded, and the moisture content of subsamples of larvae and frass were determined by oven-drying at 105 ◦C until constant weight [27]. The total dry content of larvae and frass were calculated. Subsamples of larvae were used for the protein and lipid content determination as well as enzyme activity analysis. The rest of the samples were stored at −20 ◦C for further analysis.


**Table 1.** Experimental design of nitrogen source supplied in Trials 1 and 2.

<sup>1</sup> The C and N contents were determined for food waste (FW, C 47.0%, N 2.26%, *n* = 2), soybean flour (C 48.5%, N 5.98%, *n* = 2), and fish meal (C 39.9%, N 10.5%, *n* = 2) materials. All parameters were obtained on a dry matter basis.

#### *2.3. Analysis of Larval Development and Nutrient Composition*

Larval length and weight were determined with a two-day interval over 12 days of treatment. Body length was determined in triplicates for larvae within the same box, and the body weight was measured by a combined 10-larva weight that was averaged to obtain the single weight. Larval protein and lipid contents were determined for samples collected on Day 12. To determine protein content, larvae were freeze-dried and milled, larval C and N contents were measured with the Vario EL cube elemental analyzer, and larval protein contents were calculated as the nitrogen content × 4.67 following Janssen et al. [28]. In the lipid analysis, freeze-dried and milled larvae were extracted by petroleum ether (Aladdin) twice (1:10, *w*/*v*, 48 h, 25 ◦C) in order to achieve crude lipid according to Zheng et al. [29]. After evaporating the petroleum ether, the crude lipid weights were recorded, and the lipid contents were measured as the ratio of crude lipid weight to the larval dry matter. Furthermore, the larval total protein and lipid yield in g/100 g dry waste were evaluated based on the equations below and as mentioned previously [13,30,31].

$$Protein\ yield\ \% = \frac{L \times protein\%}{W} \times 100\tag{1}$$

$$Lipid \; \text{yield} \; \%= \frac{L \times lipid \%}{W} \times 100\tag{2}$$

where L represents the total dry matter of larvae on Day 12, protein% represents larval protein content (Day 12), lipid% represents larval lipid content (Day 12), and W represents the total dry matter of food waste, including the nitrogen supplement. All parameters were obtained in grams on a dry matter basis.

#### *2.4. Assessment of the Process Efficiency*

To assess the larval treatment efficiency, the larval survival ratio (SR), waste reduction ratio (WR), larvae yield (LY), nitrogen conversion ratio (NCR), efficiency of conversion of digested feed (ECD), as well as the mass distribution pattern were evaluated by applying the aforementioned equations [14,18,30,31]:

$$\text{Lvarival ratio (SR)} \,\%\,\,=\frac{\text{Lvarva}\_{\text{end}}}{\text{Lvarva}\_{\text{b}\%}} \times 100\tag{3}$$

$$\% \text{Maste reduction ratio } (\text{WR})\,\% = \frac{\text{W} - \text{R}}{\text{W}} \times 100\tag{4}$$

$$\text{Larvae yield } (\text{LY}) \,\,\%=\frac{L}{W} \times 100\tag{5}$$

$$\text{Nitrogen conversion ratio (NCR) }\%= \frac{L \times N\%\_{lamvae}}{W \times N\%\_{waste}} \times 100\tag{6}$$

$$\text{Efficiency of conversion (ECD)} \,\%= \frac{L}{W-R} \times 100\tag{7}$$

$$\text{Mass balance} : \text{ } \text{W} = \text{R} + \text{L} + \text{M} \tag{8}$$

where Larvaebeg and Larvaeend represent the larval numbers at the beginning and end of the treatment respectively, W, R, and L represent the total dry matter of food waste including the nitrogen supplement (W), the frass residue (R), and the larvae (L) respectively, M represents the dry matter loss as a result of larval and microbial metabolism (M), N%larvae represents the nitrogen content of larvae on Day 12, and N%waste represents the nitrogen content of food waste, including the nitrogen supplement. All parameters were obtained in grams on a dry matter basis.

#### *2.5. Enzyme Activity and Amino Acid Composition Analysis*

Following C/N adjustment with urea in Trial 2, the larval enzyme activity and amino acid composition were further analyzed in triplicate. The activities of antioxidant enzymes, such as peroxidase (POD), superoxide dismutase (SOD), catalase (CAT), and glutathione peroxidase (GSH-px), were observed and recorded. Fresh larvae collected on Day 12 were homogenized by a 1× phosphate-buffered solution (PBS, pH 7.4) in a 1:10 (*w*/*v*) ratio and then centrifuged at 10,000× *g* for 3 min. The supernatants were collected and analyzed for the POD, SOD, CAT, and GSH-px activity following the practice of Chen et al. [32]. Briefly, the POD activity was determined at 420 nm using substrate containing H2O2, and 1 U of POD was defined as the amount of enzyme that catalyzed 1 μg substrate per liter per minute. The SOD activity was determined at 550 nm through the xanthine and xanthine oxidase system, and 1 U of SOD was defined as the amount of enzyme that created 50% inhibition of xanthine oxidase. The CAT activity was analyzed by measuring the absorbance decrease at 240 nm due to H2O2 decomposition, and 1 U of CAT was defined as the amount of enzyme that decomposed 1 μmol H2O2 per liter per second. The GSH-px activity was measured at 423 nm in a system containing 5,5-dithio-bis-(2-nitrobenzoic acid) (DTNB) and reduced GSH, and 1 U of GSH-px was defined as the amount of enzyme that oxidized 1 μmol of reduced GSH per liter per minute. All enzyme activities were calculated and recorded in the unit of U/g larval wet weight.

As for the amino acid composition analysis, freeze-dried and milled larval samples were oxidized by performic acid at 4 ◦C for 16 h and then hydrolyzed by 6 M HCl with 0.1% phenol at 110 ◦C for 22 h. The hydrolyzed aliquots were then diluted by 0.02 M HCl and analyzed using a L-8900 High-speed Amino Acid Analyzer (Hitachi High-Tech, Japan) following the instruction manual. For Trp (trytophan), the freeze-dried and milled

larval samples were hydrolyzed by 5 M NaOH at 110 ◦C for 22 h, and the hydrolyzed solutions were tested for Trp through a F-7000 fluorescence detector (Hitachi) with an excitation wavelength of 280 nm and an emission wavelength of 340 nm, as described by Gold et al. [30].

#### *2.6. Statistical Analyses*

All the statistical analyses were carried out using R 3.4.1 [33]. Differences among groups in the Trials 1 and 2 were tested using analysis of variance (ANOVA) provided by the *multcomp* package [34], which was associated with the *TukeyHSD* function for the pairwise comparison of means. Significance was defined as *p <* 0.05.

#### **3. Results**

#### *3.1. Effects of Nitrogen Source on the Food Waste Treatment*

Nitrogen source showed limited impacts on the larval length (*df* = 9, *F* = 2.287, *p* = 0.059) and crude protein content (*df* = 9, *F* = 1.398, *p* = 0.253), while in contrast, exerting significant effects on the larval weight (*df* = 9, *F* = 8.167, *p* < 0.001), protein yield (*df* = 9, *F* = 16.410, *p* < 0.001), crude lipid content (*df* = 9, *F* = 3.800, *p* = 0.01), and lipid yield (*df* = 9, *F* = 20.690, *p* < 0.001) as shown in the Figure 1. After 12 days of treatment, the larval body length reached 13.4–15.4 mm (Figure 1A) and crude protein content increased to 36.3–40.6%, regardless of the nitrogen source difference (Figure 1C). For the larval weight, the NaNO3 group showed significantly lighter weight than the other groups, with an average at 0.0423 g when other groups' averages ranged from 0.0688 to 0.0959 g (Figure 1B). For the protein yield, both NH4Cl (3.5%) and NaNO3 (1.1%) groups were significantly lower than that of the blank control (7.8%), while the results from other groups (7.1–9.3%) were similar to that of the control (Figure 1D). For the crude lipid content, none of the nitrogen source groups were significantly different from the blank control (25.1%), ranging from 21.3% to 30.9% (Figure 1E). For the lipid yield, results from NH4Cl (2.6%) and NaNO3 (0.8%) groups were significantly lower than that of the blank control (4.9%), with the result of the soybean flour group (8.3%) significantly higher than that of the control and results from the other groups (4.7–6.9%) similar to that of the control (Figure 1F).

Nitrogen sources substantially affected the process efficiency, as shown by the indexes of WR (*df* = 9, *F* = 90.640, *p* < 0.001), LY (*df* = 9, *F* = 25.080, *p* < 0.001), SR (*df* = 9, *F* = 15.52, *p* < 0.001), NCR (*df* = 9, *F* = 18.88, *p* < 0.001), and ECD (*df* = 9, *F* = 5.438, *p* = 0.001), respectively in the Figure 2. The NH4Cl (45.8%) and NaNO3 (12.3%) groups displayed significantly lower WR index compared to the blank control (63.1%), whereas the urea group (73.5%) exhibited significantly higher WR than the control, while results from other groups (58.8–72.0%) were similar to that of the control (Figure 2A). The LY and SR indexes exhibited similar trends, that NH4Cl (LY 8.6%, SR 42.0%) and NaNO3 (LY 3.01%, SR 29.1%) groups were significantly lower than the blank control (LY 19.3%, SR 87.3%), and the other groups (LY 17.8–25.4%, SR 81.7–99.2%) were similar to the control (Figure 2B,C). The NCR index showed that results from the NH4Cl (23.1%) and NaNO3 (7.2%) groups were significantly lower than that of the control (73.5%), while data from the urea (50.0%), uric acid (48.9%), and Glu (46.8%) groups were higher than that of the NH4Cl and NaNO3 groups, whereas lower than the control, and data from all other groups (57.7–60.7%) were similar to that of the control (Figure 2D). The ECD indexes varied between 18.9% and 38.7% (Figure 2E), and none of the nitrogen groups demonstrated significantly different ECD compared to the control (30.6%). As for the mass balance analysis, the mass distribution pattern suggested that NH4Cl and NaNO3 groups resulted in substantially less larvae yield and higher frass residues compared to the blank control, while other nitrogen groups showed approximately similar distribution patterns of larvae, frass, and metabolism mass as the control (Figure 2F).

**Figure 1.** Larval growing and nutrient properties following nitrogen source amendment in Trial 1. Error bars represent standard deviations (*n* = 3). The (**A**) larval body length and (**C**) crude protein content show no significant differences among groups. The (**B**) larval body weight (*p* < 0.001), (**D**) protein yield (*p* < 0.001), (**E**) crude lipid content (*p* = 0.010), and (**F**) lipid yield (*p* < 0.001) are different among groups, and groups with different letters represent significant difference (*p* < 0.05).

**Figure 2.** Bioconversion efficiencies following nitrogen source amendment in Trial 1. Error bars represent standard deviations (*n* = 3). The (**A**) waste reduction ratio (*p* < 0.001), (**B**) larvae yield (*p* < 0.001), (**C**) survival ratio (*p* < 0.001), (**D**) nitrogen conversion ratio (*p* < 0.001), and (**E**) efficiency of conversion (*p* = 0.001) are different among groups, and groups with different letters represent significant difference (*p* < 0.05). The (**F**) mass distribution ratio shows no significant differences between groups.

#### *3.2. Effects of C/N on the Food Waste Treatment*

The C/N of food waste did not affect larval length (*df* = 5, *F* = 0.488, *p* = 0.779); however, it did influence larval weight (*df* = 5, *F* = 5.934, *p* = 0.005), crude protein content (*df* = 5, *F* = 4.962, *p* = 0.011), protein yield (*df* = 5, *F* = 7.107, *p* = 0.003), crude lipid content (*df* = 5, *F* = 9.874, *p* = 0.001), and lipid yield (*df* = 5, *F* = 8.035, *p* = 0.002) as shown in the Figure 3. Within the 6 C/N groups, larval length ranged from 14.4 to 15.9 mm (Figure 3A). Larval weight varied between 0.074 and 0.109 g (Figure 3B), and the C/N (14:1) group (0.109 g) showed significantly higher body weight than the blank control (0.081 g). Larval crude protein ratio ranged from 35.4% to 42.0%, with none of the 5 C/N group differing from the blank control significantly (Figure 3C). Protein yield ranged from 7.3% to 11.2%, and the C/N (16:1) group (11.2%) yielded significantly higher values than that of the blank control (7.8%) (Figure 3D). Larval crude lipid content changed from 21.2% to 30.3%. None of the 5 C/N groups significantly differed from the blank control, though data from the C/N (18:1–16:1) groups were higher than that of the C/N (14:1–10:1) groups (Figure 3E). Lipid yield varied between 4.4% and 8.1%, with the C/N (18:1) (7.6%) and C/N (16:1) (8.1%) groups exhibiting significantly higher data than the blank control (4.9%) (Figure 3F) and other C/N groups (4.4–6.2%), demonstrating results similar to the control.

**Figure 3.** Larval growing and nutrient properties following C/N adjustment in Trial 2. Error bars represent standard deviations (*n* = 3). The (**A**) larval length shows no significant differences between groups. The (**B**) larval weight (*p* = 0.005), (**C**) crude protein content (*p* = 0.011), (**D**) protein yield (*p* = 0.003), (**E**) crude lipid content (*p* = 0.001), and (**F**) lipid yield (*p* = 0.002) are different among groups, and groups with different letters represent significant difference (*p* < 0.05).

The C/N of food waste greatly affected the process efficiency, as indicated by the WR (*df* = 5, *F* = 29.630, *p* < 0.001), LY (*df* = 5, *F* = 4.540, *p* = 0.015), SR (*df* = 5, *F* = 6.468, *p* = 0.004), and NCR (*df* = 5, *F* = 19.400, *p* < 0.001) indexes in the Figure 4. However, the ECD (*df* = 5, *F* = 1.583, *p* = 0.238) index was less affected. The WR indexes ranged from 61.0% to 84.8%, and the C/N (18:1) (84.8%), C/N (16:1) (77.6%), and C/N (14:1) (73.5%) groups yielded significantly higher data than that of the blank control (63.1%) (Figure 4A). The LY indexes changed from 19.3% to 26.6%, and the C/N (16:1) (26.6%) group was significantly higher than the blank control (19.3%), while the other 4 C/N groups ranged from 20.6% to 25.4% of BCRs (Figure 4B). The SR indexes changed from 61.9% to 98.0%, and none of the 5 C/N groups differed from the blank control significantly (Figure 4C). The NCR indexes showed a pattern which indicated that the C/N (18:1) (83.0%), C/N (16:1) (81.3%), and C/N (14:1) (68.1%) groups were similar to the blank control (73.5%), while the C/N (12:1) (39.9%) and C/N (10:1) (35.2%) groups were lower than the control (Figure 4D). The ECD indexes varied within 29.7% and 34.6% and none of the 5 C/N groups differed significantly from the blank control (30.6%) (Figure 4E). The mass balance analysis suggested that the C/N (18:1), C/N (16:1), and C/N (14:1) groups together resulted in relatively higher larval ratios and lower frass ratios compared to the blank control, C/N (12:1), and C/N (10:1) groups (Figure 4F).

**Figure 4.** Bioconversion efficiencies following C/N adjustment in Trial 2. Error bars represent standard deviations (*n* = 3). The (**A**) waste reduction ratio (*p* < 0.001), (**B**) larvae yield (*p* = 0.015), (**C**) survival ratio (*p* = 0.004), and (**D**) nitrogen conversion ratio (*p* < 0.001) are different among groups, and groups with different letters represent significant difference (*p* < 0.05). The (**E**) efficiency of conversion and (**F**) mass distribution ratio show no significant differences among groups.

#### *3.3. Effects of C/N on Larval Enzyme Activity and Amino Acid Composition*

Among all the four larval enzymes tested, only the SOD activity (*df* = 5, *F* = 4.561, *p* = 0.015) differed greatly between the 6 C/N groups as shown in the Table 2, while the POD (*df* = 5, *F* = 2.266, *p* = 0.114), CAT (*df* = 5, *F* = 1.710, *p* = 0.207), and GSH-px (*df* = 5, *F* = 0.816, *p* = 0.561) activities did not. For the amino acid composition analysis, the compositions of each amino acid among the 6 C/N groups were not significantly different as shown in the Table 3. The averaged proportions of each amino acid were therefore calculated, and the Glu, Ala, and Asp were found to be the top 3 most abundant amino acids.


**Table 2.** Activities of antioxidant enzyme of larvae collected on Day 12 in Trial 2.

Values are presented as mean ± standard deviation (*n* = 3). Different letters represent significant differences among column-wise groups.

**Table 3.** Proximate amino acid compositions of larvae collected on Day 12 in Trial 2.


Values are presented as g/100 g protein with mean ± standard deviation (*n* = 3).

#### **4. Discussion**

#### *4.1. Effects of Nitrogen Source on the Bioconversion Process*

Among all the 9 nitrogen sources, the NH4Cl and NaNO3 resulted in markedly adverse effects on the larval development and process efficiency compared to the 7 other organic nitrogen species, suggesting that NH4Cl and NaNO3 were probably less suitable than the organic nitrogen species in terms of facilitating larval development and waste degradation. The less efficient performance in the NH4Cl and NaNO3 conditions were probably due to several reasons, including: (1) the high amount of compounds used, (2) the toxicity generated by chloride or sodium salt, and (3) the low survival and adaptability of BSFL to these environments. In general, NH4-N and NO3-N were typical nitrogen species in the leachate and sludge biowaste. Larvae that fed on liquid leachate [24] and sewage sludge [18] have been found to be of low survival rate (30–40% mortality) and low bioconversion performance (LY 0.2–2.3%), which complied with the present results in the NH4Cl and NaNO3 conditions. However, Green et al. [23] reported that feeding BSFL with 10 mM NaNO3 solution, specifically 14 mg NO3-N/100 mL solution, facilitated the BSFL's transformation of NO3-N to NO2-N and further to NH4-N (e.g., denitrification). The current study supplied 1 g NH4-N or NO3-N/100 g dry matter to the food waste. The high amount of N element could be one of the main reasons for the negative effects of NH4Cl and NaNO3 effects on the BSFL compared to Green's study. However, as a nitrogen source used for nitrogen amendment of food waste, addition of 1 g N/100 g dry matter is a reasonable requirement that NH4-N and NO3-N might fail to address the need due to their negative effects on BSFL.

Interestingly, a subpopulation of larvae survived after the NH4Cl and NaNO3 amendments in the study. This subpopulation resulted in a similar protein and lipid body ratio when compared with the larvae which grew in the 7 other organic nitrogen conditions, suggesting that a small percentage of BSFL may gain the ability to adapt to NH4Cl and NaNO3 nutrient/environment, either through the direct incorporation of the NH4-N/NO3-N or through the indirect utilization of NH4-N/NO3-N assimilated by the in vivo or in vitro microorganisms. Barragan-Fonseca et al. [35] has pointed out that the larval protein content is regulated within narrow constraints, whereas the fat content is strongly impacted by nutrient concentration. The present results agreed with these findings that larval protein content was limitedly affected by the nitrogen environment in terms of BSFL survival; however, the larval lipid content was significantly affected by the nitrogen sources, where urea, L-Glu/L-Asp, and soybean flour were probably better nutrient sources for BSFL compared to other nitrogen species due to the relatively higher lipid content or lipid yield of BSFL.

Food waste amended with the 7 organic nitrogen sources generally resulted in neither negative nor positive effects on the BSFL performance compared to the control based on the bioconversion indexes of WR, LY, SR, and ECD. One of the possible reasons could be the inefficient supplemental ratio of the organic nitrogen sources to the food waste. According to the NCR index, all the nitrogen amendment conditions were lower than the control, suggesting that the optimum nitrogen supplying amount was not achieved, and the NCR values (46.8–60.7%) of the 7 organic nitrogen conditions were less than a previous study (66.4% ± 6.5%) conducted for mussel and bread waste treatment [13]. These results suggest that the addition amount of selected nitrogen source is very important and should be critically optimized in order to achieve high performance of waste reduction and BSFL bioconversion efficiency.

#### *4.2. Effects of C/N on the Bioconversion Process*

Although there is no well-known regulation about whether urea could be used as a food additive for the BSFL, urea is a suitable nitrogen source for investigating the C/N effects on the food waste treatment by the BSFL. The reasons are as follow: (1) Urea is a nitrogen source containing the highest N content (46.7%) and lowest C content (20.0%) amongst the 7 organic nitrogen species, which allow urea to be one of the most efficient nitrogen sources used for nitrogen content amendment while simultaneously limiting the energy/nutrient effect generated by the carbon. (2) Urea exhibited a feasible nutrient effect to BSFL according to the lipid and protein analysis in the Trial 1, (3) urea is a natural food source for BSFL as it is contained in the animal feces [19,21], and (4) urea is widely available as an artificially synthesized chemical. Therefore, urea was selected as the nitrogen source used for C/N adjustment in the Trial 2 of this study.

The larval protein and lipid data suggested that the C/N range of 18:1 to 16:1 was optimal for a high larval protein and lipid yield, while the bioconversion indexes indicated that the C/N range of 18:1 to 14:1 was highly efficient for the waste reduction (WR, 73.5–84.8%, *p* < 0.001) and larvae production (LY, 10.1–11.1%, *p* = 0.003). Therefore, supplying food waste with a moderate amount of urea (adjusting food waste C/N from 21:1 down to 18:1–14:1, especially 18:1–16:1) significantly facilitated larval development and food waste consumption; however, further addition of urea (lowering the C/N down to 12:1–10:1) would result in urea waste and even negative effects on the larval growing and bioconversion process, as indicated by the declining WR, LY, and NCR indexes.

The current results suggest that overdosing nitrogen-rich material would result in larval mortality and declined process efficiency. Similar results have also been seen in two other recent studies. Lopes et al. [14] studied recycling aquaculture waste by feeding BSFL with fish waste and bread mixture, and Ewald et al. [13] has tried to manipulate larval fatty acid composition by feeding BSFL with mussel and bread mixture. Both studies suggested that adding a moderate amount of nitrogen-rich aquaculture waste was beneficial for larval development, while too much nitrogen material may lead to negative effects such as larval

mortality and biomass loss. Based on the larvae yield, this moderate range for fish waste treatment was approximately 5–15% of fish carcasses [14], and approximately 10–20% of mussel for the mussel material [13]. Unfortunately, neither studies reported the C or N content of the diet materials. If roughly assuming the bread C and N content [36] to be 48.9% of C (50% of organic matter) and 2.95% of N (16% of crude protein), and fish C and N content to be 45% of C and 66% of N (as fish meal in this study), the 5–15% of fish waste in Lopes' study is thus equivalent to C/N of 15:1–12:1, which is close to the optimum range of 18:1–14:1 found in this study, indicating the beneficial effects of modifying C/N of food waste into this range. As for the applying amount of nitrogen amendment, urea could be more efficient than aquaculture waste, as 0.36–1.10 g urea/100 g dry matter (equivalent to 0.11–0.33% wet weight basis) used in this study resulted in comparable performance improvements compared with the 5–15% of fish waste [14] or 10–20% of mussel waste [13] used in previous studies.

Interestingly, the C/N adjustment through urea altered the larval production performance but not the crude protein content or the amino acid composition. In the fish and mussel studies [13,14], a clear trend of the positive correlations between the aquaculture materials and larval protein content was observed, although higher larval mortality occurred simultaneously with more aquaculture waste used. This finding suggests that the urea may not support the BSFL growth directly with the amino acid nutrient, whereas it may improve larval development through modulating larval metabolism, such as the SOD enzyme activity. Interestingly, the amino acid composition of BSFL is relatively stable despite wide variations of C/Ns. This could be highly due to the same nitrogen source, i.e., urea, used in the current C/N trial. In another study where BSFL fed on different substrates [18], the amino acid composition among groups varied greatly, and Tyr, Glu, and Asp were found as the top 3 amino acids for BSFL fed on human feces, and Glu, Asp, and Lys were found as the top 3 species for the food waste substrates. The current study also found Glu and Asp as the top 2 amino acids, whereas the third abundant species was Ala. Compared to the previous study [18], the Cys, Met, Thr, Ser, Gly, and Pro proportions were generally higher, and the Tyr and Lys proportions were generally lower in the larvae of current study. These results suggest that the nitrogen species may influence BSFL amino acid composition more greatly than the C/N ratios.

Altogether, the waste reduction performances of 73.5–84.8% and larvae yield of 25.3–26.6% at the C/N conditions of 18:1–14:1 in the current study are higher than many of the previous BSFL studies [3,19]. Other than the higher nutrient and digestibility of food waste used in current study, the nitrogen supplement of urea and optimal range of C/N could be two of the main contributors to the performance improvement.

#### **5. Conclusions**

Adjusting the C/N of food waste substrate is a viable method for improving the larval treatment performance. Organic nitrogen is more suitable than the NH4Cl or NaNO3 as the nitrogen amendment. Urea was a reliable and practical nitrogen source for the C/N adjustment. Addition of small amounts of urea, targeting C/N of 18:1–14:1, may significantly improve the waste reduction performance, while targeting C/N of 18:1–16:1 may substantially increase the larval protein and lipid conversion efficiency, and the BSFL amino acid composition was not affected by the C/N variation. Therefore, the current study reveals that the C/N range of 18:1–16:1 is likely the optimal condition for food waste treatment by BSFL, and the application of the current strategy may improve the food waste biodegradation and facilitate the nutrient recycling by the BSFL farming.

**Author Contributions:** Conceptualization, W.X., H.Z. and J.X.; methodology, W.X. and M.W.; validation, Y.B. and H.T.; investigation, Y.L., S.Z. and S.S.; data curation, Y.L., N.J. and M.R.; writing original draft preparation, Y.L., S.Z. and W.X.; writing—review and editing, J.X., H.Z. and W.X.; funding acquisition, W.X. All authors have read and agreed to the published version of the manuscript.

**Funding:** This study was financially supported by the Fundamental Research Funds for the Central Universities (grant number DUT21LK29, DUT20LK36), Yingkou enterprise and doctor innovation program of Yingkou Science and Technology Bureau (grant number 202005), and the National Natural Science Foundation of China (grant numbers 31670767).

**Institutional Review Board Statement:** Not applicable.

**Data Availability Statement:** The data presented in this study are openly available in FigShare at doi.10.6084/m9.figshare.14609535.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


*Article*
