Standardized ileal digestible amino acids and apparent metabolizable energy content in defatted black soldier fly larvae meal fed to broiler chickens

Standardized ileal digestibility (SID) of amino acids (AA) and apparent metabolizable energy corrected for nitrogen (AMEn) in defatted black soldier fly larvae meal (BSFLM) were determined in broiler chicks. A total of 180-d-old male broiler chicks (Ross 708) were fed a commercial broiler starter diet to day 13 of age. On day 14, birds were weighed and placed in cages (10 birds per cage; n = 6) and allocated one of two semi-purified cornstarch-based diets. The diets were nitrogen-free diet (NFD) for estimating endogenous AA losses and 20% crude protein test diet with BSFLM as the sole source of AA. All diets had 0.5% titanium dioxide (TiO2) as an indigestible marker and the ratio of cornstarch to sucrose and soy oil in the test diet was identical to NFD to calculate AMEn by difference method. Birds were given feed and water ad libitum. Excreta samples were collected on days 17–20 and ileal digesta on day 21. The SID of lysine, methionine, cysteine, threonine, isoleucine, and valine was 86.3%, 88.7%, 72.8%, 85.5%, 89.6%, and 88.6%, respectively. Apparent retention of gross energy in BSFLM was 64.5% ± 2.27% and AMEn was 2902 ± 101 kcal kg−1 dry matter. The data will aid in accurate incorporation of BSFLM in poultry feeding programs.

especially high-quality animal protein will trigger an unprecedented increase in animal production. For example, the current animal protein production will need to increase 60% or more by 2050 (FAO 2011). This increase in animal protein demand will need enormous resources, the feed being the most challenging because of the limited availability of natural resources, climate change and food-feed-fuel competition (FAO 2011). This trend has clearly demonstrated the danger of relying on a limited pool of ingredients to formulate feeds and underscored the need to characterize the nutritive value of other feedstuffs with a potential to serve as alternatives to or complementary to traditional feedstuffs.
Insects have been proposed as a high quality, efficient and sustainable alternative protein source (De Marco et al. 2015;Marono et al. 2017;Dabbou et al., 2018;Mwaniki et al. 2018;Secci et al. 2018). Using insects as a protein source can contribute to global food security via feed or as a direct food source for humans (Schader et al. 2015). For example, in Canada, $27 billion worth of food ends up in landfills or composters each year (Parizeau et al. 2015). The nutrients in the organic waste could be recycled back for animal feeding through insect rearing (Rumpold and Schluter 2013;Makkar 2017). The insect species with the highest potential for large-scale production are the black soldier fly (BSF) (Hermetia illucens), common housefly (Musca domestica), and yellow mealworm (Tenebrio molitor). Specifically, BSF larvae achieve high growth rate and excellent conversion of organic waste to produce a meal (BSFLM) with consistent amino acid concentration when raised on diverse substrates (Diener et al. 2009;Nguyen et al. 2015;Spranghers et al. 2017).
The use BSFLM as a component of diet has been reported for poultry (De Marco et al. 2015;Marono et al. 2017;Dabbou et al., 2018;Mwaniki et al. 2018;Secci et al. 2018). However, in general there is a dearth of information on digestible AA and AMEn data for BSFLM. Where data do exist, most have been reported based on apparent ileal digestibility (AID) as opposed to standardized ileal digestibility (SID) estimates (De Marco et al. 2015;Schiavone et al. 2017). It has been suggested that SID estimates should be used in formulating poultry diets because these are additive in a mixture of feedstuffs compared with AID estimates (Angkanaporn et al. 1996).
Moreover, formulating using SID of AA estimates results in diets that more closely match the birds' requirements and reduce excess nutrients (Adedokun et al. 2007;Moughan et al. 2014;Adeola et al. 2016). We recently reported AA and gross energy composition of a commercial BSFLM approved for poultry feeding in Canada (Mwaniki et al. 2018). The present study determined SID of AA and AMEn value of this sample in a broiler chickens assay.

MATERIALS AND METHODS
The experimental protocol was reviewed and approved by the University of Guelph Animal Care Committee and birds were cared for in accordance with the Canadian Council on Animal Care guidelines (CCAC 2009).

Black soldier fly larvae meal and experimental diets
Defatted BSFLM (approximately 6% crude fat as fed) was procured from a commercial manufacturer and vendor (Enterra feed Corp., Vancouver, BC, Canada). The meal is a dry, powder product derived from larvae of the black soldier fly (Hermetia illucens) reared on preconsumer recycled food collected from local farms, food processors and grocery stores. The meal is approved by the Canadian Food Inspection Agency for feeding poultry and its chemical composition was previously reported (Mwaniki et al. 2018). A nitrogen free diet (NFD , Table 1), corn starch-based diet was formulated to allow estimation of basal ileal endogenous N and AA losses for the calculation of SID of CP and AA (Adeola et al. 2016 source of AA) containing diet was designed to contain 20% CP with the ratio of corn starch to sucrose to soy oil (the sole sources of energy in NFD) maintained constant to allow determination of AMEn in feed samples using the substitution method (Woyengo et al. 2010).
All the diets contained TiO 2 (0.50%) as an indigestible marker and were fed as mash.

Birds, housing and experimental procedures
A total of 180 d old male broiler chicks (Ross 708) were placed in 12 cages and fed a commercial broiler starter crumbled diet to d 13 of age. The commercial broiler starter diet (Floradale Feed Mill Ltd., Floradale, ON, Canada) was corn, wheat and soybean meal based (3,000 kcal/kg of AME, 22% CP, 0.96% total Lys, 0.40% total Met, 0.80% TSAA, 0.53% available P, 0.97% Ca, and phytase at 500 phytase units/kg). On d 14, birds were weighed and placed in cages (10 birds/cage; n=6) and allocated to diets. The balance of chicks were transferred to Arkell general flock. Birds were given feed and water ad lib. Excreta samples were collected on d 17 to 20. On d 21, all birds were sacrificed for ileal digesta

Samples preparation and chemical analyses
Daily excreta samples were pooled for each cage and oven-dried at 60°C, whereas ileal digesta samples were freeze-dried. Samples of diets, ileal digesta and excreta were finely ground in a coffee grinder (CBG5 Smart Grind, Applica Consumer Products Inc., Shelton, CT) and thoroughly mixed for analysis. All samples were analyzed for DM, Ti and N. The samples were further analyzed as follows: diets for gross energy (GE), crude fat (CF), and AA contents; ileal digesta for AA contents; and excreta samples for GE, and CF contents. Dry matter determination was carried out according to standard procedures method 930.15 (AOAC 2005 Mills et al. (1989). Briefly, about 100 mg of each sample was digested in 4 mL of 6 N HCl for 24 h at 110°C, followed by neutralization with 4 mL of 25% (wt/vol) NaOH and cooled to room temperature. The mixture was then equalized to 50 mL volume with sodium citrate buffer (pH 2.2) and analyzed using an AA analyzer (Sykam, Germany). Samples for analysis of sulfur containing AA (Met and Cys) were subjected to performic acid oxidation prior to acid hydrolysis. Tryptophan was not determined. Titanium content was measured on a UV spectrophotometer following the method of Myers et al. (2004). The NDF and ADF contents were determined according to Van Soest et al. (1991) using Ankom 200 Fiber Analyzer (Ankom Technology, Fairport, NY). For BSFLM samples, the amount of protein linked to acid detergent fiber (ADF) was determined and was used to estimate the amount of chitin, according to Marono et al. (2015), Chitin (%) = ash free ADF (%) -ADF-linked protein (%). The ADF-linked protein was derived from the concentration of N in ADF residue.

Calculations and statistical analysis
The AID, SID of CP and AA and apparent retention (AR) of components in experimental diets and BSFLM sample were calculated according to Adeola et al. (2016). The SID content of CP and AA content for BSFLM were calculated using the following equation  The AME content (kcal/kg) = [(AR of GE for BSFLM, %) × (GE content in BSFLM, kcal kg -1 )]/100. The AMEn content for BSFLM was calculated from AME as described by Woyengo et al. (2010) using the following equation: AMEn (kcal kg -1 ) = AME − (8.22 × ANR), where ANR = apparent N retained (g kg -1 of feed intake).
The AA and GE content in BSFLM were from Mwaniki et al. (2018).
Data were reported as average and standard deviation (SD).  (Mwaniki et al. 2018). The concentration of NDF and ADF in BSFLM sample was 38.7 and 12.6% DM, respectively. Insects have been reported to contain variable and significant amounts of fiber measured as crude fiber, NDF and ADF (Barker et al. 1998;Marono et al. 2015). Analyses of six

Results and discussion
Hermetia illucens samples showed concentration ranging from 6.1 to 36.3% DM NDF and 4.7 to 9.3% DM ADF (Marono et al. 2015). In plant-based feedstuffs, NDF is composed of cellulose, lignin, and hemicelluloses (Van Soest et al. 1991). Although, insects contain significant amounts of both ADF and NDF, the mono sugar components that make up these fractions are largely unknown (Finke 2007). Finke (2007) showed a significant amount of amino acids (9.3 to 32.7% by weight) associated with ADF fractions of several insect species. As ADF is part of NDF, the amount of protein linked to ADF were included both in the CP and NDF; for this reason, the sum of ash + CP + crude fat + NDF was shown in several insect samples to be higher than 130/100 g The analyzed chemical composition of the NFD and BSFLM containing diets is shown in Table 2. The AID and SID of CP in BSFLM was 80.7 and 84.6%, respectively (Table 3). Among the indispensable AA, Arg had the highest AID (88.7%) and His had the least AID (54.6%). The AID for Lys and Met were higher than for the whole BSFLM sample (46 and 42%, respectively) (De Marco et al. 2015) but comparable with values reported for defatted BSFLM (80 and 81%, respectively) (Schiavone et al. 2017). The crude fat content of BSFLM fed in the present study was 7.01% DM in previous evaluation (Mwaniki et al. 2018) and was lower than values of 15-35% DM reported for non-defatted (whole) BFLSM (Makkar et al. 2014;De Marco et al. 2015) F o r R e v i e w O n l y but comparable to defatted BSFLM samples (Marono et al. 2017;Schiavone et al. 2017).
Defatting has been shown to increase concentration of crude protein from 40-44% DM in whole BSFLM (Makkar et al. 2014;Spranghers et al. 2017) to a high of 65.5 % DM (Schiavone et al. 2017). The concentration of CP of BSFLM sample tested in the present study was 57.5% DM (Mwaniki et al. 2018). Based on the present observations and previous reports it can be interpreted that defatted BSFLM has high concentration of CP than non-defatted samples.
The AR of crude fat was 93.1% (Table 4) and was comparable to values of more than 90% reported for non-defatted and defatted BSFLM samples fed to broiler chickens and quails  Marco et al. 2015;Cullere et al. 2017;Schiavone et al. 2017). The AR of CP, NDF and, ADF were 47.0, 44.3 and 28.5 %, respectively. Moderate and variable retention of protein and energy in broilers fed BSFLM has been attributed to the negative effects of chitin (De Marco et al. 2015;Schiavone et al. 2017). Broiler chickens fed chitin derived from crustacean shell waste (37.3% CP) digested approximately 50% of chitin protein (Hossain and Blair 2007). Marono et al. (2015) demonstrated that in vitro CP digestibility was negatively correlated to the chitin content. Surprisingly, chickens have been shown to produce chitinase in the proventriculus and hepatocytes (Suzuki et al. 2002). The AR of GE in BSFLM was 64.5 ± 2.27% and the AMEn was 2,902 ± 101 kcal kg -1 DM ( Table 4) 55.3% DM CP) and 2,354 kcal kg -1 DM (4.6% DM crude fat; 65.5% DM CP) (Schiavone et al., 2017). This suggested importance of crude fat concentration on assigning accurate AMEn values of BSFLM in practical poultry feed formulation.

Conclusions
Successful application of BSFLM in poultry rations will depend on expanding digestible nutrients and energy database to document variability in poultry. The present data generated information to allow accurate incorporation of BSFLM in practical poultry diets. The BSFLM