Impact of chick mash, rice bran, and wheat bran as starter feeds on the performance of black soldier fly larvae in food waste treatment

Acronym

Meaning

BSFL

black soldier fly larvae

FW

food waste

WRE

waste reduction efficiency

BCR

bioconversion rate

FCR

feed conversion rate

RB

rice bran

WB

wheat bran

CM

chick mash

Meeting the food demand of the ever-growing human population, while reducing the negative environmental impacts of agricultural production systems, is one of the greatest challenges of the 21st century (Foley et al. 2011). The adverse effects of food production on the environment are mainly a result of excessive fertilizer application and food wastage, which pose a threat to human and ecosystem health through the enrichment of water bodies, causing eutrophication (Grizzetti et al. 2013). Eutrophication's negative effects on aquatic ecosystems include the spread of harmful algal blooms, biodiversity and habitat loss, water hypoxia, and fish kills (Bonsdorff 2021). In addition, nutrient enrichment of the water bodies results in the reduction of their aesthetic value, which impairs their use for drinking purposes. As such, of key interest in this study is the management of food waste (FW) generated at the consumption level.

FW, according to Parfitt et al. (2010), is ‘wholesome edible material intended for human consumption that arises at any point of the food supply chain but is instead discarded, lost, degraded, or consumed by pests.’ The food supply chain includes many processes, from harvesting to transport, storage, processing, packaging, distribution, and consumption. According to the Food and Agriculture Organization (FAO) statistics, about one-third of the food produced globally for human consumption is lost or wasted through the steps of the supply chain, with higher per capita values of FW in Europe and North America (95–115 kg/yr) and lower values in Sub-Saharan Africa and South/Southeast Asia (6–11 kg/yr) (Gustavsson et al. 2013). Global assessments have drawn attention to the negative effects of increasing losses of nitrogen and phosphorous from FW to the environment, emphasizing how the excessive input of nutrients into the environment poses a threat to human and aquatic life and leads to significant economic loss (Chapagain & James 2013; Grizzetti et al. 2013). Therefore, reducing FW could be one of the measures for improving nutrient use efficiency and reducing nutrient loss in the environment.

The use of insects such as black soldier fly larvae (BSFL) has recently gained a lot of attention in the bioconversion of organic wastes (Laganaro et al. 2021; Matheka et al. 2021). Insect farming is both economically and environmentally viable because it utilizes biochemical products and by-products such as egg proteins and lipids. Furthermore, the required insect breeding space is not large compared with the large land area required for crops or the large water footprint required for microalgae production (Hasnol et al. 2020). The larvae have strong mouthparts and powerful digestive enzymes that enable them to effectively breakdown various organic wastes (Cho et al. 2020), reducing unpleasant smells and undesirable bacteria by modification of the bacterial microflora by BSFL during feeding (Singh et al. 2021). The BSFL feed on a variety of organic substrates, including FW, fecal waste, poultry manure, cattle manure, sewage sludge, and abattoir waste (Lalander et al. 2019; Kim et al. 2021; Matheka et al. 2021).

The use of BSFL in organic waste management keeps the carbon footprint low and assists in recycling refuse. In addition, BSFL is not a vector that can transmit diseases or parasites when feeding; it is therefore neither harmful to animals nor humans (Diener et al. 2011). According to Diener et al. (2011), the use of BSFL in organic waste treatment has the potential to protect the environment from nutrient enrichment of aquatic systems, odor, and greenhouse emissions from composting organic wastes, and increase food security by providing animal feeds.

Despite the benefits of BSFL in waste treatment, their survival and optimal growth are dependent on suitable conditions such as feed components, adequate temperature, humidity, and acidity (Meneguz et al. 2018). The nutrient composition of feed components (biowaste) has been established to have the largest influence on the performance of the larvae (Nguyen et al. 2013). Proteins, non-fiber carbohydrates, and lipids are highly digestible by BSFL, and therefore, their supply enhances performance (Gold et al. 2019). Conversely, feeds containing fiber, cellulose, and lignin are less digestible and tend to decrease larval growth rates (Liu et al. 2018). Also, the ash content of biowaste negatively correlates with bioconversion rates (BCRs) (Lalander et al. 2019). However, among the essential nutrients required for optimal BSFL growth, several studies have concluded that the protein content of biowaste is most important (Liu et al. 2018; Kim et al. 2021). Hence, an insufficient amount of proteins may prolong development, reduce growth and related biomass production, and limit the efficiency of waste reduction (Gold et al. 2019).

To date, many studies have reported on the potential of BSFL for treatment of biowaste, with specific focus on determining the optimal feeding rates, optimal larval weight gain, BCRs, waste reduction efficiency (WRE), protein conversion efficiency, and survival rates (Sheppard et al. 2002; Diener & Studt 2011; Banks et al. 2014; Nyakeri et al. 2017). Despite a variety of substrates having been investigated for rearing BSFL, there is limited information on the impact of starter feeds on the performance of BSFL in FW treatment. This is important in selecting the most suitable starter feed to enhance the BCR in FW treatment. Starter feeds are mainly used at the initial stages of larval development, immediately after egg hatching, after which they are transferred to the waste substrate. This study therefore focused on the impact of starter feeds (chick mash (CM), rice bran (RB), and wheat bran (WB)) on the performance of BSFL in FW treatment, with specific objectives encompassing: (i) analysing the nutrient composition of the starter feeds; (ii) determining the bioconversion efficiency, WRE, and survival rate of BSFL fed on the starter feeds; and (iii) modeling the growth kinetics of BSFL in FW treatment using the Gompertz model.

This study was a laboratory-based experiment in a greenhouse located at Jomo Kenyatta University of Agriculture and Technology (JKUAT) in Juja, Kiambu County, Kenya, at longitude 37° 7′ 0″ E and latitude 1° 11′ 0″ S. The temperature and relative humidity of the greenhouse ranged between 28–38 °C and 40–60%, respectively.

RB and WB were collected from the Mwea Rice Mill, while CM was obtained from chicken feed stores in Thika, both in Kenya. FW, mostly composed of rice and beans waste, was collected from a restaurant at JKUAT. The BSF eggs were obtained from the Sanergy Ltd production site and transported to the study site for hatching.

The proximate parameters (moisture, carbohydrate, lipids (fats), protein, and ash) of the substrates were determined according to the procedures recommended by the Association of Official Analytical Chemists (AOAC 1984) as briefly described in the following.

High performance liquid chromatography (HPLC) with a refractive index detector was used for the determination of total sugar content as detailed in a study by Zielinski et al. (2014). The volatile fatty acids (VFAs) were determined using the gas chromatography method as detailed by Baawain et al. (2017). The total organic carbon (TOC) was determined using high-temperature combustion at 1,200 °C in an oxygen-rich atmosphere and the CO₂ produced was passed through scrubber tubes to remove interferences and measured by non-dispersive infrared absorption. A similar procedure for the determination of carbon content in municipal solid waste was used by Ferrari et al. (2002). For the volatile solids (VSs), gravimetric analysis method was adopted as described by Baawain et al. (2017).

The carbon and nitrogen content of the samples was determined using the Thermo Flash 2000 CHNS/O elemental analyser.

Precisely 500 g of the dry weight of CM, WB, and RB were accurately weighed and transferred to clean plastic containers. The starter feeds were each mixed with 300 g of water such that the water constituted 60% (by weight) of the dry feed weight according to Diener et al. (2009). Approximately 1 g of eggs was weighed and put on each substrate in the plastic containers. Similarly, 1 g of eggs was put on 500 g of FW with a moisture content of 65.5 wt.%. The hatchlings were observed after 3–4 days in the starter feed and in the FW. The newly hatched larvae were allowed to feed on the starter feeds and the FW until they were 5 days old. On the sixth day, the larvae were sieved through a 1.5-mm-diameter mesh screen to obtain larvae of approximately the same size and weight for each treatment. The sampling was done from among the sieved larvae by counting 300 larvae from each treatment. A pair of forceps was used to handle the delicate larvae. The 300 larvae from each treatment were then weighed on an electronic weighing scale to determine the average collective weight and, consequently, the mean initial weight per larva from each treatment. Figure 1 represents sampling of starter feeds while Figure 2 represents BSFL being fed on FW.

The 300 sampled larvae from each treatment (CM, WB, RB, and FW) were divided into three groups, each comprising 100 larvae, and then transferred to the FW in feeding trays of dimensions 260 × 130 × 110 mm. The larva transferred onto the feeding trays had an average weight of 0.0031, 0.0026, 0.0015, and 0.0021 g/larva, from the CM, WB, RB, and FW, respectively. Overall, 12 treatments resulted since the three groups of larvae fed on each of the three starter feeds and FW were transferred to the main FW to be treated. Using a uniform feeding rate of 100 mg/larva/day of wet waste as recommended by Diener et al. (2009), the substrates were weighed and distributed to the feeding containers. Each larval group was then evenly spread on the allocated substrate surface. The feeding was continued for a total of 16 days, by which time most larvae changed into black/brown prepupae (Stratiomyidae et al. 2017). Every second day, 50 larvae were randomly counted from each feeding tray and collectively weighed. The total weight obtained was divided by the number of larvae to obtain an average weight for each of the larvae. The weight gains after every 2 days were calculated by comparing the obtained average larval weight with the previous mean larval weight. All the prepupae and any larvae were then harvested by sieving them through a 5-mm-diameter mesh screen and the residue material dried in an oven at 105 °C for 24 h to determine its dry mass. The final prepupae weight was taken to be the average weight of the biomass recorded on the day of harvest.

The constants P_o, R_max, and t_o were fixed by the Origin Pro nonlinear fitting program. The BSFL growth data that were used in modeling were continuous data as from the first day of transfer from starter feeds to the main substrate (FW) up to the 30th day. Similarly, the BSFL growth entirely in the FW (control) was monitored for the same duration of time. Based on the findings by Nyakeri et al. (2017), BSFL development takes 16 days beyond which no weight gain occurs since the insect enters the prepupae stage where no active feeding takes place. Hence, the data set was sufficient to simulate the growth of the BSFL in FW treatment. All the treatments were done in triplicates in the same period, and the means of the larval weight gain measured after every two days for the duration of the experiment were used to simulate the growth of BSFL using the Gompertz model.

A two-tailed t-test at 95% confidence level was used to determine if there was significant difference in the means of optimal larval weight, waste reduction, BCRs, and FCRs from the treatments.

Table 1 shows the mean nutrient composition of the starter feeds and FW. The protein content ranged from 12.59 to 17.68% with CM having the highest protein content. Lipid content was highest in FW (17.8%) and lowest in WB (4.89%). Ash content was highest in RB (21.94%), while being lowest in FW (3.98%). Carbohydrate content ranged from 45.37 to 58.75% with RB having the lowest carbohydrate content. Other parameters investigated (VFA, total sugars, nitrogen, TOC, carbon-to-nitrogen (C/N) ratio, and VS) also varied among the substrates.

It can be deduced that two distinct growth stages of BSF were experienced during the duration of monitoring the experiment. The phase in which rapid weight gain occurred up to an optimum (0–16 days) could be associated with the larval development stage when active feeding occurs, while the phase in which no weight gain occurred (>16 days) could be associated with the prepupae stage when the BSF stop feeding. In a similar study whereby the BSFL was reared on household food waste, Broeckx et al. (2021) observed that the larval weight of BSFL increased rapidly and reached a maximum on the 8th day beyond which a decrease in mean larval weight was observed. The authors attributed the decline in weight to the transition of the larvae to the prepupae stage in which the larvae stop feeding and spend energy to either pupate or search for a suitable place to pupate. Also, Nyakeri et al. (2017) observed that the mean BSFL weight increased rapidly in food waste between 0 and 16 days, which is consistent with our findings. The study by Nyakeri et al. (2017) did not however monitor BSFL growth beyond 16 days to capture the optimal weight gain reached before prepupae stage begins. According to Kinasih et al. (2018), the variation in duration of BSFL development in organic wastes could be related to the balance in nutritional contents, with the nutritional imbalance in diet leading to an increase in consumption period of insect larvae to compensate for deficiency in the nutrients especially proteins and carbohydrates.

The better performance of BSFL fed on CM and WB starter feeds before transfer to the FW and also BSFL feed entirely on FW could be attributed to their protein and nitrogen content, which are easy digestible by the larvae during early stages of growth (Kinasih et al. 2018; Liu et al. 2018; Broeckx et al. 2021; Singh et al. 2021). On the contrary, despite RB having protein content comparable with other substrates (Table 1), the lower performance of BSFL fed on RB starter feed could be attributed to its higher ash content, which impairs the digestive process of BSFL, thereby negatively correlating with BCRs (Meneguz et al. 2018; Lalander et al. 2019; Broeckx et al. 2021). Overall, in the selection of starter feeds, the focus should not only be on the quantity of digestible nutrients but also on the amount of indigestible ash fractions, which negate BSFL growth. Thus, nutritious starter feeds with lower ash content should be selected for rearing BSFL.

The performance of BSFL in consuming and consequently reducing the FW is presented in Table 2.

The efficiency of BSFL in consuming organic wastes is determined by calculating waste reduction, BCR, and FCRs (Banks et al. 2014). The BSFL that were initially fed on CM, significantly (P < 0.05) consumed the highest amount of FW compared with other treatments, thereby reducing the waste by 82.8%, while the larvae that were transferred from RB consumed the lowest amount of FW, thereby reducing the waste by 75.5%. This concurs with the observation presented in Figure 3, in which larvae fed CM attained the highest optimal larval weight in FW, while those fed RB attained the lowest optimal larval weight in FW.

Nevertheless, the substrate reduction in the range of 75.5–82.8% obtained in this study is comparable with the value of 81.8% food waste reduction by BSFL reported by Nyakeri et al. (2017). Interestingly, the percentages of material reduction by BSFL obtained in the present study and by Nyakeri et al. (2017) are higher than that reported for chicken manure of 50%, municipal organic waste of 68%, and fresh human feces of 46% (Craig Sheppard et al. 1994; Diener & Studt 2011; Banks et al. 2014). The inherent substrate composition and structure influences the BSFL development (Banks 2014; Kim et al. 2021); hence, the variation in substrate reduction between the food wastes and other organic wastes. For instance, proteins, non-fiber carbohydrates, and lipids are highly digestible by BSFL, and therefore, their supply enhances performance and, consequently, substrate reduction (Gold et al. 2019). Conversely, feeds containing fiber, cellulose, and lignin are less digestible and tend to decrease larval growth rates (Liu et al. 2018). Although feeds containing fiber, cellulose, and lignin are less digestible, the ability of BSFL to reduce such waste is due to the ability of BSFL to secrete enzymes and harbor microbes that can degrade plant materials (Kalová & Borkovcová 2013).

The BCR is normally an indication of the effectiveness of the larvae to reduce organic waste, and the higher the value, the higher the effectiveness of the larvae in waste reduction (Diener & Studt 2011). The larvae transferred from the CM to FW exhibited significantly (P < 0.05) higher BCR compared with other treatments with larvae transferred from WB to FW and those fed entirely on FW having significantly (P < 0.05) similar BCR. The RB-fed larvae depicted significantly (P < 0.05) the lowest BCR in FW. BCR ranging from 8.1 to 13.8 were observed in this study for the treatment of the FW substrates regardless of starter feed type, as shown in Table 3. These values were comparable with the 11.8 reported for the treatment of municipal organic waste (Diener & Studt 2011), but lower than the 22.3 reported for the treatment of fresh human feces (Banks et al. 2014) and higher than the 3.7 reported for the treatment of chicken manure (Craig Sheppard et al. 1994). The BCR reported in this study showed the effectiveness of the larvae to consume the substrates, which is an indicator that the FW had a high nutritive value and that the substrate was not only effectively consumed but also highly assimilated into the larval biomass. It is worth mentioning the higher BCR of 20.8 reported for restaurant FW by Nyakeri et al. (2017). The FW used by Nyakeri et al. (2017) was highly heterogeneous, composed of meat pieces, rice, beans, and so on, as opposed to the FW used in this study, which was largely rice and beans. Hence, the variation in the level of protein content in the FW could have resulted in the difference in BCR.

Conversely, although FCR is also an indicator of the effectiveness of the larvae to reduce organic waste, in contrast to BCR, a higher FCR value means that less larval weight gain occurs in a given amount of substrate used for rearing the larvae, which consequently indicates that the substrate is of less nutritional value and thus largely excreted (Diener & Studt 2011; Banks et al. 2014). Similar trend to waste reduction and BCR was observed in FCR values in this study with larvae transferred from CM to FW portraying significantly (P < 0.05) the lowest FCR value compared with other treatment. The larvae transferred from CM to FW showed significantly (P < 0.05) the highest FCR value, while larvae transferred from WB to FW and those fed entirely on FW showed significantly (P < 0.05) similar FCR values. Generally, FCRs ranging from 7.3 to 12.3 were observed in this study for the treatment of FW substrates regardless of starter feed type, as shown in Table 3. These values are however higher than the FCR value of 2.6 reported for treatment of restaurant food waste (Nyakeri et al. 2017) as well as a value of 5.8 reported by Diener et al. (2011) for the treatment of municipal organic waste. The values reported in this study are however comparable with the FCR value of 13.4 reported for treatment of chicken manure (Craig Sheppard et al. 1994).

Overall, the BCR and FCR values obtained in this study are an indication that BSFL could effectively be used to manage FW by converting them into organic fertilizers in the form of excreta and nutritious larval biomass.

The key factors that influence larval growth and development time are feed availability, nutrient availability, and feed characteristics (Diener & Studt 2011; Lalander et al. 2019). The steep slope of Figure 4(a), which depicts BSFL growth in FW after initial treatment in CM, indicates that the larvae used less time to achieve their maximum larval weight when compared with those of other treatments. Moreover, the BSFL fed on CM starter feed attained the highest weight gain, an indication of higher assimilation of nutrients for growth. Nevertheless, the P < 0.05 in all the set-ups (Table 2) indicates that all the treatments affected the BSFL weight, which shows that the larvae can still degrade FW even if starter feeds are not used, although starter feeds of high nutrient content could enhance the optimal weight gain and consequently waste reduction.

Despite using the same FW to feed BSFL, different P_o values were obtained, which is an indication that BSFL activity was different in all the treatments. Higher BSFL activity generally indicates the affinity of the larvae for the substrate thereby resulting in higher FCR (Li et al. 2015). The BSFL activity in T1 was the highest with T3 exhibiting the lowest performance. It can thus be stated that the type of starter feed used influences the feeding characteristics of the BSF larvae in the main substrate. As such, feeding the BSFL in a more nutritious starter feed enhances their early development thus enabling them to consume more food as they grow toward maturity. It is worth mentioning that the higher P_o and BSFL activity obtained in the control where no starter feeds were utilized (T4) than when RB was used (T3) is a further indication that starter feed contents, especially ash, could affect the conversion rate of BSFL in the degradation of FW. A similar trend was exhibited by the maximum conversion rate (R_max), whereby maximum conversion occurred in T1 and the minimum in T3. Similarly, the R_max in the control (T4) of 0.025 was higher than 0.015 when RB was used as a starter feed (T3). However, the coefficient of correlation R² > 0.95 in all the treatments indicates that the model was adequate in predicting and describing the BSFL growth in FW.

The purpose of this study was to determine the effect of CM, WB, and RB as starter feeds on the performance of BSFL in FW treatment. The use of starter feeds influenced the performance of BSFL in the treatment of FW since BSFL transferred from CM and WB to the FW showed significantly (P < 0.05) higher waste reduction of 82.8 and 80%, respectively, compared with the 75.5% obtained when BSFL was fed on RB starter feeds. The BCR depicted that feeding the BSFL on CM starter feed resulted in significantly (P < 0.05) highest effectiveness in degradation of the FW (13.8%) followed by larvae fed on WB starter feed (11.3%) and then those fed on RB starter feed (8.1%). A similar trend was observed by the feed conversion ratio (FCR) that portrayed CM to significantly (P < 0.05) enhance the assimilation of nutrients and conversion from FW to the larval biomass, followed by the use of WB and RB as starter feeds, respectively. The better performance of larvae transferred from CM to FW was attributed to its high protein content that enhanced early development of larvae while the low performance of larvae transferred from RB to FW was attributed to their higher ash content, which impaired larvae digestive process during their development stage. The modified Gompertz model adequately fitted the experimental data (R² > 0.95) in all the treatments, thereby reliably predicting and describing the growth of BSFL in the FW. On the basis of the above findings, the study recommends the use of CM and WB as starter feeds for use in the farming of BSFL to enhance their conversion rate in FW, thereby supplying the much-needed protein alternative and organic waste bioremediation.

This work was supported by the African Development Bank (AfDB) in partnership with the Ministry of Education and the Jomo Kenyatta University of Agriculture and Technology.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.