Despite the losses and damages caused by FAW, it acts as a food source to some birds and grasshoppers which are also sources of food to human beings (SANBI, 2018). Furthermore, materials from the environment (insects, worms, snails, greens, seeds among others) are possible feed sources for indigenous chicken (FAO, 2013). Besides, insects that are considered as pests in agro-ecosystems can also be used as food or feed (Dobermann et al.,  2017; van Huis, 2016, van Huis and Oonincx, 2017).Therefore, FAW larvae could be used as an alternative source of protein for indigenous chicken.

Mechanical control is one of the methods of controlling the fall armyworm in crop fields (Prasanna et al., 2018). Consequently, farmers have started to collect the FAW larvae by handpicking and feeding to the chicken C. Midega, (Personal communication, April, 2019) instead of crushing them. Therefore, FAW larvae could be a promising source of protein. However, there is little information on its nutritional profile. Information on the nutrient quality of FAW larvae would enable people know the benefits of the insect for feed, hence increasing their use.

Indigenous chicken are commonly found in the traditional system of poultry production in the developing countries. The chickens are scavengers which require feeds that can provide them with the necessary nutrients for increased production of eggs and meat. This is because they get very low nutrient inputs during scavenging (Atela et al., 2016). Harvesting of FAW larvae from the crops field and feeding to these chickens could possibly boost their nutrition.

Insects are sources of very high-quality feed (Abro et al., 2020; Duinkerken et al., 2012) that can be supplemented in poultry diets. Black soldier fly and the common housefly can be very good sources of protein for poultry. The black soldier fly larvae /BSFL) has been reported to contain 36.6 - 62.7% crude protein and 14.0 - 40.7% fat (Abro et al., 2020; Ebeneezar et al., 2021; Ewald et al., 2020; Duinkerken et al., 2012; Mohammed et al., 2017; Shumo et al., 2019).

The common housefly (Musca domestica) which has been utilized mostly as poultry feed is reported to constitute between 43 – 68% crude protein and 4 – 32% fat (Elahi et al., 2020; Fitches et al., 2019; Duinkerken et al., 2012; Hussein et al., 2017; Pieterse and Pretorius, 2014). According to Duinkerken et al. (2012), the crude protein content in common housefly is comparable to soybean meal which is a conventional poultry feed.

Previous studies have shown that fall armyworm fed on artificial diet had crude protein of 59% and fat of 20.6% while those fed on fresh plant materials had a crude protein of 59.3% and fat of 11.7% (Williams et al., 2016). Therefore, FAW from crop fields could act as an alternative source of protein. The aim of the study was to compare the nutritional profile of FAW larvae to other chicken feeds in order to provide information on the quality of FAW larvae as poultry feed. As  a  result,  more people would use FAW larvae as feed, FAW populations will reduce, and maize yields would increase thus a boost to food security.

Sample collection

The samples were collected from two sites. The first site was at the School of Agricultural and Food Sciences fields (Formerly Farmers Training Centre (FTC)) located in Siaya County at Siaya town in Kenya, East Africa. The county lies approximately between latitude 0º 26´ South to 0º 18´ North and longitude 33º 58´ and 34º 33´ East (Latitude.tob, 2021). The samples were collected during the short rains of 2020.

Since the study could not get enough FAW larvae for three samples during the short rains of 2020, more FAW larvae were collected in the long rains of 2020 from a maize field in Sindo sub county of Homabay County. Homa Bay County lies between Latitude: 0° 31' 38.32" N and Longitude: 34° 27' 25.70" E (Latitude.toc, 2021).

Method of sample collection

S. frugiperda samples were collected by hand picking the larvae from the maize plants in the field (that is collected from the wild). This was done by observing the maize plants with signs of FAW infestations and then the larvae were removed from the whorl of the plant carefully. The larvae were put together as one composite sample once the study was certain it was sufficient to be used for a complete proximate analysis.

The samples were S. frugiperda (S1), S. frugiperda (S2) and S. frugiperda (S3), where S1 (mature larvae 5th instar (L5) or 6th instar (L6)) were collected from maize fields in Siaya Campus – Siaya County during the short rains of 2020. The S2 and S3 were various stages of the larvae ranging from 2nd instar (L2) to 6th instar (L6), which were collected from a maize field in Homa bay County during the long rains of 2021.

Sample preparation

Fresh samples were left for 24 h to degut and then blanched in boiling water (Ayieko et al., 2016)for about 5 min. Thereafter, they were left in the sun to dry for at least 8 h before they were sent to the laboratories for nutritional analysis. S. frugiperda S1 was sent to Kenya Agricultural and Livestock Research Organization (KALRO) in Kakamega County Kenya. The other two samples S. frugiperda S2 and S. frugiperda S3 were analysed at the Animal Sciences Laboratory in Egerton University, Njoro, Kenya. The samples were taken to Egerton University for analysis after the machines in KALRO broke down.

Proximate analysis

The samples at KALRO were analysed in duplicate, where crude protein analysis was done using a standard Kjeldahl method and a neutral detergent fibre analysis (Zaklouta et al., 2011)was performed for crude fiber. At Egerton University Laboratories, air-dried samples were heated in an oven at 60°C for 2 h. The samples were then ground through a 2-mm screen and kept in an air-tight bottle until analysis (Fombong et al., 2017). The analysis for moisture, crude protein, crude fiber and ash were done in triplicate while that for fat was done in duplicate.

Moisture content was determined by oven drying at 105°C for 8 h. The loss in weight was the moisture content and what was left was the dry matter of the sample. Crude ash was obtained through incineration of the sample in a furnace at 550°C for 4 h, while the crude protein was determined by micro Kjeldahl method, where the protein content = N x 6.25 (conversion factor). The crude fats or ether extract (EE) were determined using Soxhlet extractor method (Ayieko et al., 2016; Zaklouta et al., 2011).

The crude fibre was determined as described by Nduko et al. (2018) by weighing about 2.000 g of air-dried sample into a 600-ml glass beaker in triplicate, where 100 ml of hot water was added before adding 2.04 N H₂SO₄ and then the volume was increased to 200 ml. The content of the beaker was boiled for 30 min but the level of the solvent was kept at 200 ml by adding hot water. Thereafter, the beakers were removed and the content filtered using a filter stick packed with glass wool.

The residue was washed 3 times using hot water and then returned into the beakers into which 100 ml of hot water was added followed by 25 ml of 1.78 N KOH, then the volume was increased to 200 ml using hot water in order to keep it constant. This was boiled for 30 min after which it was filtered and washed 3 times using hot water. The residue was transferred into crucibles and then dried in an oven set at 105°C for 2 h, cooled in a desiccator and weighed accurately. The contents were then burnt in a furnace at 550°C to ash for 4 h and then left to cool to about 100°C. After which the sample were transferred to a desiccator for further cooling to room temperature and then their weights taken immediately.

Carbohydrate content (that is, Nitrogen free extracts) of the insects were determined by subtracting the sum of the weights of protein, fiber, lipid and ash from the total dry matter weight (Fombong et al., 2017).

Comparison of the nutrient content of S. frugiperda larvae to other chicken feeds

The primary data on proximate analysis of S. frugiperda larvae from the current research was compared to secondary data of the Black soldier fly, the common housefly and soya bean. The study chose to make comparison to Black soldier fly (BSF) larvae and the common housefly (HF) larvae because they have been widely accepted as poultry feed. The study only considered BSFL that were fed on food waste namely: Household waste, kitchen waste and departmental canteen waste. Soya bean data was also used in the study because of its high use as feed for many livestock (Abro et al., 2020; Allegretti et al., 2017).

Statistical analysis

The data was visualized and analyzed using R software for statistical analysis R version 4.0.5 (2021-03-31) -- "Shake and Throw". The distribution of the FAW samples data was checked using Shapiro-Wilk normality test while homogeneity of variance was performed using Bartlett’s test for dry matter, moisture, fat, crude fibre and carbohydrates. Levene's Test was used to check crude protein and Ash. Kruskal-Wallis rank sum test was used to analyse the FAW samples data for dry matter, moisture, ash, crude protein and fat; and then Dunn’s Kruskal–Wallis Multiple Comparisons (Dunn’s Test) was used to perform the post hoc test where there was significant difference.

Mann Whitney U Test (Wilcoxon Rank Sum Test) was then used to compare the differences between S2 and S3 FAW samples for crude fibre and carbohydrates. It was also used to test the differences between: FAW larvae and BSF larvae, FAW larvae and HF larvae and finally FAW larvae and Soya bean.

The results are presented in figures and tables showing visualization and analysis of the data for the proximate composition of the fall armyworm samples, black soldier fly, the common house fly and soya bean.

Proximate composition of FAW larvae

Figure 1 shows the proximate composition of the three samples of S. frugiperda (S1, S2 and S3). It indicates that the samples had almost similar amounts of ash, crude fibre, ether extract and carbohydrate content except for the crude protein. S1 was different in the crude protein content from S2 at p = 0.04  but similar to S3 (Table 1).

Proximate composition of BSF larvae and FAW larvae

The study revealed that FAW larvae has higher amount of crude protein (%) than BSF larvae but low amount of fat compared to BSF larvae (Figure 2). However, there were no statistical differences at p = 0.05 between the crude protein (%) of FAW larvae and BSF larvae. The fat content also differed at p = 0.04 for FAW larvae and BSF larvae (Table 2).

Proximate composition of housefly (HF) larvae and FAW larvae

The results in Figure 3 depict that FAW larvae and the common housefly (HF) larvae have the same amount of nutrients content and they are rich in crude protein. This can also be seen in Table 3 where there are no significant differences in the nutrient contents of FAW larvae and BSF larvae.

Proximate composition of Soya bean and FAW larvae

Visualization of the proximate composition of FAW larvae and Soya bean (Figure 4) showed that FAW larvae had higher crude protein content than soya bean. Nevertheless, the results of the crude protein were not statistically significantly different from each other (Table 4).

In the current study, there was significant difference in the crude protein between S. frugiperda (S1) and S. frugiperda (S2) at p = 0.04 (Table 1). This is consistent with the report by Rumpold and Schl (2013) that there are differences  in  the  results  of  nutritional  composition   of edible insects. In this study, the difference could have resulted from the method that was used for the analysis. S. frugiperda (S1) was analyzed at KALRO while S. frugiperda (S2) and S. frugiperda (S3) were analysed at Egerton University. The difference could also result from the stage of the larvae that was harvested; for S. frugiperda (S1) it was L5/L6, while the larval stages of S. frugiperda (S2) and S. frugiperda (S3) ranged from L2 to L6. Therefore, the study suggests that further research be conducted on the nutritional profile of different larval stages.

The percent crude protein of S. frugiperda was comparable to that of Spodoptera littoralis 51.2% (Sayed et al., 2019). Our findings are in agreement with those reported by Williams et al. (2016) that  the  percent  crude protein of S. frugiperda fed on artificial diet and S. frugiperda fed on fresh plant materials were 59.0 and 59.3% respectively. When the crude protein of FAW larvae was compared to BSF larvae (Table 2) and housefly (HF) larvae (Table 3) and Soya bean (Table 4), there were no statistical differences in their means. These results indicate that FAW larvae is rich in protein. Besides, caterpillars in the order lepidoptera have been reported by several authors to be rich in protein (Braide et al., 2010; Sayed et al., 2019; Williams et al., 2016).

There were no statistically significant differences in the amount of crude fibre available in S. frugiperda (S2) and S. frugiperda (S3) (Figure 1). A comparison of the crude fibre of FAW larvae, BSF larvae, HF larvae and Soya bean  indicated  that  the  amount  of  crude  fibre was the same (Tables 2 to 4). Crude fibre of < 10% for S. frugiperda in the current study is in line with the report of Williams et al.  (2016),  which  is  an  indication  that  their bodies were not hardened.

The average ash content did not have any significant differences among the S. frugiperda samples in this study (Table 1). This was similar to the report by Williams et al. (2016) of 5.7 for S. frugiperda fed on artificial diet but they also reported a slightly higher value of 11.6 for S. frugiperda larvae fed on fresh plant products. According to Braide et al. (2010), ash content of 6.42% is an indication of high mineral content. Therefore, this research also found the ash content of 5.6 and 7.4, hence S. frugiperda is rich in minerals.

A comparison of FAW larvae ash content to HF larvae (Table 3) and Soya bean (Table 4) did not have statistical significant differences. Therefore, this shows that FAW larvae possibly have the same mineral content as HF larvae and Soya bean. Nevertheless, there was statistical significant difference when it was compared to BSF larvae (Table 2). This indicates that BSF larvae could be having higher amount of minerals compared to FAW larvae.

There was no significant difference in the fat content of S. frugiperda S1 (22.9%), S2 (17.8%) and S3 (21.1%) (Table 1). The results are consistent with the reports by Williams et al. (2016) that S. frugiperda fed on artificial diets had 20.6% fat but lower in S. frugiperda fed on fresh plants 11.7%. A comparison of FAW larvae fat content (19.66%) to HF larvae (15.33%) (Table 3) and Soya bean (20.33%) (Table 4) showed that the fat content was the same but there was a statistical significant difference when the fat content of FAW larvae was compared to BSF larvae (37.3) (Table 2) (p = 0.04) which can also be seen in Figure 2. These findings are in agreement with the report by Kou?imská and Adámková (2016) who opined that averagely edible insects contain about 10 to 60% of fat in dry matter.

Insects have been reported to be having very low levels of carbohydrates (Fombong et al., 2017). This study reports carbohydrates of (1.65 – 3.39%) for S. frugiperda S2 and S. frugiperda S3 (Table 1) which are close to 5.2 for S. littoralis (Sayed et al., 2019). Since FAW larvae is comparable to BSF larvae, HF larvae and Soya bean in terms of crude protein and crude fiber except for fat and ash content with BSF larvae, there is sufficient evidence the FAW larvae could be a potential source of poultry feed. However, there is need to study if regions could possibly have an effect on their nutrient content.

Utilizing FAW larvae as a source of protein could contribute to improved global food security through feed, new business ventures, create new jobs, improve income by providing a cheaper source of poultry feed and reducing food - feed competition (Abro et al., 2020; Veldkamp and Bosch, 2015).

Based on the findings of this study, it is evident that S. frugiperda has sufficient nutrients for chicken feed. Therefore, there is a possibility that FAW larvae could be utilized as an alternative source of protein for the indigenous   chicken  feed,  since  FAW  is  rich  in  crude protein at (36.9 - 63.54%), crude fibre (9.1 - 9.6%) and crude fat (17.8 - 22.9%). Even though S. frugiperda has depicted a potential to be utilized as poultry feed due to its rich nutrients, there is still need to do further analysis to determine its safety as food or feed.

The authors have not declared any conflict of interests.