Inclusion of Dried Black Soldier Fly Larvae in Free-Range Laying Hen Diets: Effects on Production Efficiency, Feed Safety, Blood Metabolites, and Hen Health

Abstract

Identifying alternative feedstuffs to replace conventional nutrient sources in poultry diets is crucial to supplying the growing demand for animal feed. A 17-week-long feeding experiment with three diets, including 0% (control), 10%, and 18% full-fat dried black soldier fly larvae (DBSFL), was conducted to evaluate the production efficiency and feed safety of using the larvae for partial (50%) and full (100%) substitutions of soybean meal and 90% replacement of soybean oil in free-range laying hen diets. Thirty hens (18–36 weeks old) were housed in two mobile poultry trailers per treatment level. The weight gain of hens, their feed intake, egg production, egg weights, feed conversion ratios, bird welfare parameters, hematology and blood metabolites, fecal microbiology, and digestive tract weights were examined. Control hens had higher weight gains, laid more and bigger eggs while consuming less feed, and had lower feed conversion ratios than 18% DBSFL hens. However, the production performances of 10% DBSFL hens were not significantly different from the control in many of the parameters considered (e.g., hen-day egg production or HDEP). In conclusion, partial replacement of soybean meal and oil with DBSFL in layer diets maintains production efficiency, feed safety, and hen health and welfare status.

1. Introduction

Soybean meal and oil are excellent sources of protein and metabolizable energy (respectively) in laying hen diets, and they have been widely used as feed ingredients in commercial poultry production for several decades. The consumer demand for animal products (in particular poultry products) has shown significant growth in the last few decades. The latest projections indicate that global demand for food in general will increase at least by 50–60% from 2019 to 2050 [1]. This demonstrates that the need for animal feed will also increase. Soybean cultivation for poultry feed requires large areas of arable land that otherwise can be used to produce plant-based sources of food for direct human consumption. Thus, it is important to identify extra feed sources, using alternative nutrient sources that are not used directly for human consumption, to supply the growing demand for animal feed [2].

Animal processing by-products (e.g., meat meal and blood meal) and fish meal have also been used as sources of protein in poultry feed for decades [3]. However, the use of animal by-products is severely restricted in some countries (e.g., the European Union [4] because of public health concerns related to their use [5]). Furthermore, the use of animal by-products is not allowed in organic production systems (e.g., the National Standard of Canada [6], indicating that organic poultry farmers also face a scarcity of protein sources for organic poultry diets [7]. In addition, fish meal prices have increased significantly over time due to static production and a growing demand from the aquaculture industry [8].

The need for alternative feed sources in animal production has facilitated the development of circular food systems in which food waste, by-products, and inedible nutrients for humans can be used to produce feed for animals [2]. When the reduction of food loss and waste is not feasible, reusing the lost and wasted nutrients by recycling them through circular food systems is of great importance. Nevertheless, there are diverse standards and guidelines in different regions concerning which lost or wasted food products can be recycled back directly into the food system through animal feed. About one-third (31%) of the food produced in the world is lost or wasted, and approximately half of this wastage (14%) happens at the retail and consumer levels [8]. At this stage, the nutrients are still of high quality and not contaminated. Insects and worms have shown potential for recovering nutrients from food waste by establishing a natural recycling and recovery system for the production of highly nutritious feedstuffs. Their production is considered a viable solution for tackling both the increased demand for animal feed and the challenges caused by the production of excessive amounts of food waste in our current food systems [2,3,9,10].

Many alternative nutrient sources have been tested in the last few decades to identify novel feed ingredients that can improve the environmental, economic, and social sustainability of animal farming while increasing feed production. For example, the use of insects in aquaculture, poultry, and swine production has been authorized in several countries/regions [11]. Nevertheless, there are considerable differences in the insect production systems and conditions in different regions (e.g., what is allowed to be used as the nutrient source for insects). These differences have made comparison of published results difficult (see discussion) [9,10,12].

The black soldier fly (Hermetia illucens) is one of the insects used to convert significant amounts of organic waste into biomass rich in protein and lipid [3,13,14]. The use of black soldier fly larvae (from now on referred to as BSFL) has been promising in aquaculture and monogastric animal production as a substitute for conventional lipid and protein sources in aquaculture and poultry feed [2,10,14,15,16,17,18,19,20,21,22]. The nutritional value of BSFL as a feedstuff depends on the nutrient sources provided to the insect and the production stage of BSFL [21,23]. Different studies have used BSFL in different formats (full fat, defatted, chopped, grounded, meal, etc.) at different stages of animal production. Moreover, potential pathogen contents of the larvae (e.g., Escherichia coli O157:H7 and Salmonella spp. and Enterococcus spp.) and the presence of heavy metals (such as As, Cd, Cr, and Pb) or other chemical contamination are concerns about their use as a feed ingredient [24,25,26], but investigations have rarely been reported in the published literature.

Considering the knowledge gaps in this area, this study was conducted to investigate the effects of the inclusion of full-fat dried BSFL (from here on referred to as DBSFL), either as partial or full replacements of soybean meal and oil, on the production efficiency (i.e., weight gain of hens, their feed intake, egg production, egg weight and feed conversion ratio), bird health and welfare parameters, hematology and blood metabolites, fecal microbiology and digestive tract length, and weight of commercial laying hens, from the beginning of their egg production until their peak egg production. The DBSFL used in the current study was produced from pre-consumption plant-based food waste.

2. Materials and Methods

The current experiment was conducted under the guidelines of the Canadian Council on Animal Care, and the study was certified by the University of British Columbia (UBC) Animal Care Committee (Certificate #A16-0227).

2.1. Source of DBSFL

Whole DBSFL provided by Enterra Feed Corporation (Langley, BC, Canada) was used in this study as a protein and lipid source in the laying hen diets in the form of chopped larvae (about 3–5 mm length). The larvae were produced from pre-consumption plant-based food waste that was mainly supplied by the food processing and retail sectors. The larvae were harvested as a 4th or 5th instar prior to darkening into prepupae to minimize chitin content. Further information about the production practices was considered proprietary and was not shared by the producer.

2.2. Experimental Design

The feed trial was conducted at the Kwantlen Polytechnic University Teaching and Education Centre (Richmond, BC, Canada), and the laboratory work was conducted at the UBC Avian Research Centre (Vancouver, BC, Canada).

A Completely Randomized Design (CRD) study with 3 dietary treatment levels (0%, 10%, and 18% DBSFL) and 2 replications per diet was carried out for 17 weeks, starting when the hens were 18 weeks old.

Ninety commercial brown layer pullets (Novogen BROWN) were obtained from a local supplier. Each experimental unit, consisting of 15 layers, was housed in a 3.35 m (L) × 1.5 m (W) × 1.8 m (H) rat and predator-proof mobile poultry trailer, custom-built according to Canadian General Standards Board guidelines for Organic production systems [6]. The outside free-range area provided 0.33 m² per hen, and the off-the-ground area inside the hut area provided 0.17 m² per hen. Each of the six trailers was equipped with four 35.7 cm (L) × 33.0 cm (W) × 35.7 cm (H) nest boxes, inside perches (~20 cm/hen), two feeders (one inside and one outside under cover), a 13.25 liter water bucket with four nipple drinkers, and a sandbox for dust bathing. Trailers provided an enclosed environment for hens by covering all areas with wire mesh except for the ones made of wood (Figure S1). All six trailers were located in a field seeded with rye grass and white clover. Trailers were moved to new ground once every two weeks so that all birds could have access to fresh cover crops at the same time. A solar-powered electric fence was installed all around the field to minimize human and animal disturbances. From here on, the three dietary treatment groups will be referred to as 0% DBSFL (or control), 10% DBSFL, and 18% DBSFL, respectively.

2.3. Diet Formulation and Feed Nutrient Analysis

Before formulating the diets (Table 1), the proximate compositions of the DBSFL used in this study were analyzed. Proximate composition analysis indicated that its moisture, dry matter, crude protein, crude fiber, crude fat, and crude ash contents were 4.2%, 95.8%, 37.9%, 7.4%, 43.6%, and 6.4%, respectively. The ME value of DBSFL was calculated at 5.2 kcal/g based on the Carpenter and Clegg equation (−53 + 38 (% crude protein + 2.25 × % ether extract + 1.1 × % starch + % sugar)) [27]. This information was used to formulate the three experimental diets using the WUFFDA program (University of Georgia Extension, Athens, GA, 2014). The diets were formulated to be isocaloric and isonitrogenous according to NRC [28] laying hen nutrition requirements. In the partial replacement diet, 50% of the soybean meal was removed and 10% DBSFL could be included, along with some modifications in the inclusion rates of other feedstuffs to balance the diet to the same protein and energy levels as the control diet. In the full replacement diet, all the soybean meal was removed, and 18% DBSFL could be included with some modifications in the inclusion rates of other ingredients to balance the diet to the same energy and protein levels. All three diets met the requirements for essential amino acids and fatty acids and were prepared by a local organic feed supplier (In Season Farms, Abbotsford, BC). Feed was provided in the mash form in all three diets, and whole DBSFL were chopped (size of 3–5 mm) before their inclusion in the treatment diets.

In the 10% DBSFL and 18% DBSFL diets, DBSFL substituted 50% and 100% of the soybean meal, respectively. Moreover, in both diets, DBSFL also replaced 90% of the soybean oil because larvae have a high lipid content. The chemical composition, fatty acid profile, 12 amino acid concentrations, and metal profile of the three experimental diets were analyzed based on the standards defined by AOAC [29].

Table 1. Ingredients and calculated composition of three experimental diets ¹.

Ingredients (% w/w)	Diets
	0% DBSFL 2	10% DBSFL	18% DBSFL
Chopped DBSFL	0	10	18
Soybean meal	20	10	0
Sunflower meal	14.3	12.3	15
Soybean oil	4.95	0.5	0.5
Ground Corn	49.5	41.5	30
Barley	0	3.16	10
Wheat	0	9	4
Wheat bran	0	1	10
Dehydrated alfalfa meal	1	3	3.4
DL-Methionine	0.15	0	0
Limestone	8.65	8.12	7.7
Dicalcium phosphate	0.5	0.5	0.5
Common Salt	0.35	0.32	0.3
Vitamin Mineral Premix	0.4	0.4	0.4
Choline	0.1	0.1	0.1
Phytase Enzyme	0.1	0.1	0.1
Calculated composition
Dry matter (%)	91.13	91.24	91.83
M.E. (Kcal/g)	2.79	2.79	2.79
Crude protein (%)	17.02	17.02	17.02
Ether extract (%)	7.9	6.73	9.9
Calcium (%)	3.52	3.53	3.53
Na (%)	0.18	0.18	0.18
Total phosphorus (%)	0.54	0.51	0.59

¹ Feedstuff nutrient values provided by NRC [28] and Feed Supplier (In Season Farms), and laying hen nutrient requirements provided by [28,30] considering 120 g feed/hen/day; 2 DBSFL: Dried black soldier fly larvae.

Table 1. Ingredients and calculated composition of three experimental diets ¹.

Ingredients (% w/w)	Diets
	0% DBSFL 2	10% DBSFL	18% DBSFL
Chopped DBSFL	0	10	18
Soybean meal	20	10	0
Sunflower meal	14.3	12.3	15
Soybean oil	4.95	0.5	0.5
Ground Corn	49.5	41.5	30
Barley	0	3.16	10
Wheat	0	9	4
Wheat bran	0	1	10
Dehydrated alfalfa meal	1	3	3.4
DL-Methionine	0.15	0	0
Limestone	8.65	8.12	7.7
Dicalcium phosphate	0.5	0.5	0.5
Common Salt	0.35	0.32	0.3
Vitamin Mineral Premix	0.4	0.4	0.4
Choline	0.1	0.1	0.1
Phytase Enzyme	0.1	0.1	0.1
Calculated composition
Dry matter (%)	91.13	91.24	91.83
M.E. (Kcal/g)	2.79	2.79	2.79
Crude protein (%)	17.02	17.02	17.02
Ether extract (%)	7.9	6.73	9.9
Calcium (%)	3.52	3.53	3.53
Na (%)	0.18	0.18	0.18
Total phosphorus (%)	0.54	0.51	0.59

¹ Feedstuff nutrient values provided by NRC [28] and Feed Supplier (In Season Farms), and laying hen nutrient requirements provided by [28,30] considering 120 g feed/hen/day; 2 DBSFL: Dried black soldier fly larvae.

2.4. Experimental Protocols and Production Performance Data

Pullets arrived as 18-week-old birds in the last week of June. The first two weeks were an adaptation period, and the birds were fed the control diet ad libitum. All birds were weighed and tagged with numbered wing tags at the end of the first week. After the second week, all the birds were gradually introduced to their respective diet treatments. By the end of the third week, the transition and adaptation stages were completed. The data collection started at the beginning of Week 4, when the birds were 22 weeks old, and continued until Week 17 (Week 36). This study focused on the early egg production stage because it can have a long-lasting effect on production efficiency and hen health. During this stage, hens not only initiate laying eggs, reach their highest laying rate, and complete a 10-week peak production, but also gain considerable body weight [31]. Individual hen weights were recorded biweekly during the experiment period (total of 7 times).

Each trailer was provided daily with 2.2 kg of feed evenly distributed between the two feeders, and the leftover feed was weighed weekly using a handheld digital scale (Salter-Brecknell SA3N253, ElectroSamson Digital Scale, Fairmont, MN). Then, the daily feed intake per hen and feed conversion ratio (FCR) were calculated.

The number of eggs produced in each trailer was recorded daily, and then the weekly hen-day egg production (HDEP) was calculated.

Daily egg production in each trailer was weighed once a week using a pocket scale (100 g with 0.01 g reliability), and the results were extrapolated to represent the mean weekly egg weight per trailer.

2.5. Blood Chemistry and White Blood Cells (WBC) Differential Counts

Baseline blood samples (1 mL) were collected from three randomly selected hens in each trailer from the brachial vein using a tuberculin syringe with a 25 G heparinized needle on Week 2 before the hens were put on the experimental diets. The second blood sample was collected in Week 17 from the same birds immediately after decapitation (for post-mortem examination) by collecting trunk blood into heparinized vacutainer tubes. Blood was stored in BD Vacutainer^® Heparin Tubes (light green-topped tubes, Franklin Lakes, NJ, USA) and transported to the laboratory immediately after.

Blood biochemistry metabolites were analyzed to determine the concentrations of albumin, albumin/globulin ratio, amylase, aspartate transaminase enzyme, bile acids, blood urea nitrogen, calcium, chloride, cholesterol, creatine kinase, creatinine, gamma-glutamyl transferase enzyme, globulin, glucose, phosphorus, potassium, sodium, sodium/potassium ratio, total protein, triglycerides, and uric acid. Blood smears were also made for the WBC differential counts. Heterophil: Lymphocyte (H:L) ratios were calculated by dividing the absolute heterophil count by absolute lymphocyte counts. Moreover, packed cell volume (PCV or hematocrit) was measured after centrifuging blood in microhematocrit tubes at 12,000× g for 5 min.

2.6. Fecal Microbial Analysis

Fecal culture examination was conducted at Week 16, close to the end of the 17-week experiment, to study the pathogen content of hens and monitor the impact of the inclusion of the novel feed ingredient on hen health. This was carried out to ensure that the hens were exposed to the new ingredient for the longest time possible in this experiment and to detect any potential influence of the use of DBSFL on hen pathogen contents. About 5 g of fresh fecal samples were collected from each trailer by placing five randomly selected birds from each trailer in a clean crate with a plastic bag under the crate for half an hour. For each trailer, three fecal samples were pooled in a ProtocolTM C&S Medium container (Cary and Blair Medium containing 15 mL of Phenol Red solution; Fishe Diagnostics, VA, USA) (in total, six samples) for the detection of enteric pathogens. Standard bacterial growth media were used for specific incubation times and temperatures determined by the standard procedures applied by IDEXX Laboratory (Delta, BC, Canada) for the detection of enteric pathogens.

2.7. Health and Welfare Parameters of Laying Hens

Feather and wound scores [32] were recorded biweekly when the hens were weighed. Plumage conditions in six body parts (neck, breast, cloaca/vent, back, wings, and tail), bumble foot lesions, and pecking damage to the comb and the skin of the rear body were assigned scores of 1 to 4 (4 being the best integument conditions and 1 being the worst) [32].

2.8. Post-Mortem Examination of the Gastrointestinal Tract

At the end of the feed trial (Week 17), the three hens per trailer (in total, 18 hens) that were sampled in the baseline sampling were transported to the Necropsy Laboratory of the UBC Centre for Comparative Medicine. A post-mortem examination was performed to assess the effects of the three diets on the gastrointestinal tract weight of laying hens. The digestive tract of the hens was segmented into the proventriculus, gizzard (ventriculus), small intestine (duodenum, jejunum, and ileum), ceca, large intestine, liver, and pancreas [33]. The segments were weighed using a precision electric balance (0.01 g reliability). The proportional weights of the digestive tract segments were calculated using the segment weight divided by the bird’s live weight taken just before decapitation. The individual segment weights divided by the total intestinal tract weight were also calculated as the digestive tract weight proportions.

2.9. Statistical Analysis

CRD was considered the general design of the experiment, with two replications (trailers) and a single-fixed effect treatment (with three levels: 0%, 10%, and 18% DBSFL). Trailers were assigned randomly to the treatment levels.

The experiment was designed to benefit from the strengths of the repeated measures mixed models as discussed by Schank and Koehnle [34] and Olejnik and Algina [35]. The application of repeated measures tests makes it possible to obtain unbiased results with limited samples [34]. These tests increase the precision of measurements, result in a reduction of measurement errors, and increase the likelihood of identifying statistically significant differences [34,35].

The α-level was 0.05 in all tests. The data were analyzed by Least Squares ANOVA with repeated measures using JMP^® PRO software (version 17.0.0, SAS Institute Inc., Cary, NC, USA). Tukey HSD post hoc tests were used for means separation depending on the frequency of measurements, the presence of subsamples, and the type of analysis. Trailers were considered the experimental units, and the statistical models listed below were used in the present study to determine differences between treatment groups.

• CRD with repeated subsampling was used when measurements were conducted repeatedly during the experiment on hens (i.e., subsamples) within each trainer (i.e., experimental unit), e.g., blood biochemistry, WBC differential, and hen weight.
• CRD repeated in time were used when measurements were conducted repeatedly on the trailers without any subsampling, e.g., feed intake, HDEP, egg weight, and FCR.

Thus, the power of the tests was calculated for the CRD and repeated in time to check if the sample size was adequate for the tests. Cohen [36] recommended 0.8 as the generally accepted minimum level of power for tests.

Due to the limited sample size, statistical analyses were not performed on the feed composition and post-mortem data. However, the data are presented because they can provide valuable insights and trends that can be used to inform future research.

3. Results

3.1. Feed Composition Analyses

Proximate composition of the control, 10% DBSFL, and 18% DBSFL diets indicated their moisture (9.3%, 10.2%, 9.8%), dry matter (90.7%, 89.8%, 90.2%), crude protein (20.1%, 18.5%, 19.9%), crude fiber (7.3%, 7.2%, 9.1%), and crude fat (9.4%, 8.1%, 12.6%) contents (respectively). The crude fat concentration was higher in the 18% DBSFL diet because its lipid content was higher in absolute terms compared to the control and 10% DBSFL diets.

3.1.1. Metal Profiles of Diets

Table 2. Metal profiles of three experimental diets.

Metal (ppm on Dry Matter Basis)	Diets
	0% DBSFL	10% DBSFL 1	18% DBSFL
Aluminum	134.53	186.13	180.43
Antimony	<5.00	<5.00	<5.00
Arsenic	<2.50	<2.50	<2.50
Barium	3.97	4.78	6.51
Boron	11.48	7.56	5.96
Cadmium	<0.50	<0.50	<0.50
Calcium	41,087.08	48,163.13	44,186.61
Chromium	7.17	7.22	6.07
Cobalt	<1.00	<1.00	<1.00
Copper	25.49	26.34	20.29
Iron	269.29	300.03	261.39
Lead	<2.50	<2.50	<2.50
Magnesium	1941.29	2052.45	2546.60
Manganese	162.56	154.13	181.21
Mercury	<10.00	<10.00	<10.00
Molybdenum	0.99	1.00	0.88
Phosphorus	5439.80	5080.56	6159.70
Potassium	9695.40	8287.59	8569.54
Selenium	<10.00	<10.00	<10.00
Sodium	2285.62	1984.66	1701.78
Sulfur	2808.74	2450.27	2421.97
Thallium	<10.00	<10.00	<10.00
Zinc	156.16	160.13	161.46

¹ DBSFL: Dried black soldier fly larvae.

shows that similar mineral mixes were included in the three diets, in approximately similar concentrations. The Ca concentrations were higher than the recommended level in all three experimental diets used in this study. It was higher in the DBSFL diets compared to that of the control feed. The 18% DBSFL diet had the lowest concentrations of Cu and Cr compared to the other two experimental diets. Zn concentrations were similar in three diets. The As, Cd, and Pb concentrations of the three diets were negligible.

3.1.2. Amino Acid Profiles of Diets

Figure 1 shows the amino acid content of three experimental diets. Threonine, valine, isoleucine, and lysine concentrations were higher in the three experimental diets than the recommended concentrations by Novogen [30]. Methionine concentrations of DBSFL diets were slightly lower than those of the control diet and the recommended concentration in the laying hen diets [30].

3.1.3. Fatty Acids Profiles of Diets

The results of the fatty acid analysis indicated that the lipid (ether extract) contents of the three diets seemed to differ from each other (Figure 2). The lipid content of the 0%, 10%, and 18% DBSFL diets was 8.48%, 7.27%, and 11.3% (as fed), or 9.35%, 8.1%, and 12.6% (dry matter), respectively. The lauric acid and myristic acid contents of both DBSFL diets were higher than those of the control feed. In addition to that, the palmitic acid and oleic acid contents of the 18% DBSFL diet were higher than those of the control group. On the other hand, linoleic acid and linolenic acid levels were higher in the control feed than those in the DBSFL diets.

3.2. Hen Weight

There was a significant (p = 0.014) Time × Diets interaction. The weight gains of hens from Week 4 to Week 16 differed depending on the level of DBSFL in the diet (Figure 3). Results indicated that the control hens gained more weight at a faster rate during the experiment compared to the weight gains of the hens in the 10% and 18% DBSFL diets. However, the live weight of 10% DBSFL hens was not significantly different from that of the 18% DBSFL hens throughout the trial, until the last weighing (p < 0.05).

3.3. Feed Intake

There was a significant (p = 0.04) Time × Diet interaction effect (

Table 3. Fixed effect tests and interaction effects Least Square Means (LS Means) of Time × Diet on the feed intake, hen-day egg production (HDEP), egg weight, and feed conversion ratio (FCR) from Week 4 (age 22-week-old) until Week 16 of the experiment fed three diets with different levels of dried black soldier fly larvae (DBSFL). (The means followed by different letters were significantly different at p < 0.05 and n = 78.)

		Feed Intake		HDEP		Egg Weight		FCR
		F Ratio	p-Value (Power)	F Ratio	p-Value (Power)	F Ratio	p-Value (Power)	F Ratio	p-Value (Power)
Fixed Effect Tests	Treatment	12.48	0.04 (1.00)	1.75	0.31 (1.00)	73.99	<0.01 (1.00)	17.31	0.02 (1.00)
	Time	2.73	0.01 (0.94)	2.88	0.01 (0.95)	17.52	<0.0001 (1.00)	1.42	0.20 (0.81)
	Interaction	1.92	0.04 (0.93)	4.19	<0.0001 (1.00)	1.82	0.05 (0.91)	4.64	<0.0001 (1.00)
Diets	Experi. weeks	LS Mean	Sig. level	LS Mean	Sig. level	LS Mean	Sig. level	LS Mean	Sig. level
0% DBSFL	4	123.33	ABCDEF	0.98	ABCDE	56.89	HIJK	2.21	ABCDEFGH
	5	123.33	ABCDEF	0.97	ABCDE	59.52	CDEFGHIJ	2.13	CDEFGHI
	6	122.33	ABCDEF	0.97	ABCDE	60.70	ABCDEFG	2.07	DEFGHI
	7	120.33	CDEF	0.99	ABCD	62.28	ABCDEF	1.95	GHI
	8	121.67	BCDEF	0.97	ABCDE	62.50	ABCD	2.01	EFGHI
	9	121.33	BCDEF	0.98	ABCDE	63.05	AB	1.96	GHI
	10	118.67	EF	0.99	ABCDE	63.12	A	1.91	GHI
	11	120.00	CDEF	0.97	ABCDE	62.31	ABCDE	1.99	FGHI
	12	118.33	EF	0.97	ABCDE	62.58	ABC	1.96	GHI
	13	114.67	F	1.00	ABC	62.86	ABC	1.82	I
	14	126.00	ABCDEF	0.99	ABCD	63.16	A	2.01	EFGHI
	15	127.00	ABCDEF	0.98	ABCDE	62.55	ABC	2.07	DEFGHI
	16	119.33	DEF	1.00	ABC	63.22	A	1.89	HI
10% DBSFL	4	126.67	ABCDEF	0.98	ABCDE	55.43	K	2.34	ABCDEFGH
	5	125.33	ABCDEF	0.97	ABCDE	56.18	IJK	2.30	ABCDEFGHI
	6	124.00	ABCDEF	0.94	ABCDE	57.09	HIJK	2.30	ABCDEFGH
	7	127.67	ABCDEF	0.97	ABCDE	59.49	CDEFGHIJ	2.22	ABCDEFGHI
	8	134.00	ABCDEF	0.97	ABCDE	58.77	EFGHIJK	2.35	ABCDEFGH
	9	134.00	ABCDEF	0.98	ABCDE	59.92	ABCDEFGH	2.28	ABCDEFGHI
	10	127.00	ABCDEF	0.97	ABCDE	58.98	DEFGHIJ	2.23	ABCDEFGHI
	11	127.67	ABCDEF	0.99	ABCDE	60.17	ABCDEFGH	2.15	CDEFGHI
	12	124.67	ABCDEF	0.95	ABCDE	59.93	ABCDEFGH	2.19	BCDEFGHI
	13	127.33	ABCDEF	0.99	ABCDE	60.22	ABCDEFGH	2.15	CDEFGHI
	14	130.00	ABCDEF	0.97	ABCDE	59.50	BCDEFGHIJ	2.25	ABCDEFGHI
	15	131.67	ABCDEF	0.97	ABCDE	60.12	ABCDEFGH	2.25	ABCDEFGHI
	16	129.67	ABCDEF	0.97	ABCDE	60.38	ABCDEFGH	2.22	ABCDEFGHI
18% DBSFL	4	125.33	BCDEF	0.98	ABC	55.99	JK	2.28	BCDEFGHI
	5	131.67	ABCDEF	1.01	A	57.43	GHIJK	2.27	CDEFGHI
	6	129.67	ABCDEF	1.00	AB	57.32	GHIJK	2.27	CDEFGHI
	7	132.00	ABCDEF	0.97	ABCD	57.59	GHIJK	2.36	ABCDEFGH
	8	136.00	ABCDE	0.98	ABC	58.74	FGHIJK	2.36	ABCDEFG
	9	141.33	AB	0.98	ABCD	58.13	GHIJK	2.49	ABCD
	10	137.33	ABCDE	0.97	ABCDE	58.01	GHIJK	2.45	ABCDEF
	11	135.00	ABCDE	0.95	ABCDE	57.46	GHIJK	2.47	ABCDE
	12	142.33	A	0.91	DE	58.63	GHIJK	2.67	A
	13	139.00	ABCD	0.93	BCDE	58.02	GHIJK	2.58	ABC
	14	140.00	ABC	0.91	CDE	58.35	GHIJK	2.63	AB
	15	139.00	ABCD	0.94	BCDE	58.03	GHIJK	2.55	ABC
	16	142.33	A	0.90	E	59.51	BCDEFGHI	2.66	A
SE		3.38		0.02		0.60		0.08

). Hens from the 18% DBSFL dietary treatment level consumed more feed compared to hens from the other two groups. The differences in the feed intake of the 18% DBSFL hens and the control hens were significant at Weeks 12, 13, and 16 (p < 0.05).

3.4. Hen-Day Egg Production (HDEP)

There was a significant (p < 0.0001) Time × Diet interaction effect on HDEP (Table 3). The results indicated that control and 10% DBSFL hens maintained high HDEP during the study (Weeks 4 to 16), but the HDEP of 18% DBSFL hens started to decrease at Week 11 and had lower HDEP than the control hens in the last 4 weeks of the study (p < 0.05), even though they still maintained above 90% HDEP.

3.5. Egg Weight

The mean egg weights from the 0%, 10%, and 18% DBSFL diets during the experiment were 61.90 ± 0.24 g, 58.93 ± 0.24 g, and 57.94 ± 0.24 g, respectively. The average weights of the eggs from all three diets were within the weight range for large eggs in the Canadian egg grading system (56 g to 63 g). There was a significant (p < 0.05) Time × Diet interaction (Table 3) in the egg weight variable. In Weeks 4 and 5 of the feed trial, eggs produced by hens in the three treatment groups were not significantly different in weight (p < 0.05). After Week 6, the egg weight of the 10% DBSFL diet started to show some inconsistent pattern in that they were significantly lighter than those from the control diet in Weeks 6, 8, 10, and 14 (p < 0.05) but not significantly different at other time points. The egg weights of the 18% DBSFL diet were significantly lighter than control eggs after Week 6 (p < 0.05). There were no significant differences in the egg weights of the 10% and 18% DBSFL diets throughout the study.

3.6. Feed Conversion Ratio (FCR)

Weekly FCR was significantly (p < 0.0001) affected by Time × Diet interaction (Table 3). Control hens had the best FCR throughout the study. Starting at Week 9, the FCR of 18% DBSFL hens was significantly higher than the FCR of the control hens (p < 0.05). Except for the first 3 weeks, the FCR of 10% DBSFL hens was between the FCRs of the control and the 18% DBSFL hens. The 10% DBSFL hens’ FCR was significantly better than the 18% DBSFL hens at Week 12 (p < 0.05). Both 0% and 10% DBSFL hens maintained their FCR throughout the study, but the FCR of 18% DBSFL hens gradually increased.

3.7. Blood Metabolites

3.7.1. Blood Biochemistry

The baseline blood samples were taken during the ready-to-lay stage (20 weeks old), and as a result, changes in the blood parameters after 15 weeks of laying eggs were expected (

Table 4. There was a significant difference in plasma parameters between 20-week-old hen baseline plasma samples and 35-week-old hen blood samples (36 samples within 12 experimental units).

Plasma Parameters	Age		SEM	F Ratio	p-Value
	20 Weeks	35 Weeks
Glucose (mmol/L)	4.27	9.73	0.46	27.73	0.01
Urea (Bun) (mmol/L)	0.32	0.68	0.04	20.69	0.01
Uric acid (mmol/L)	437.06	149.5	21.05	176.07	0.001
Calcium (mmol/L)	6.28	4.77	0.15	57.64	0.005
Sodium/Potassium ratio	39.50	26.17	0.67	640.00	0.0001
Chloride (mmol/L)	118.56	124.17	0.86	37.09	0.01
Total protein (g/L)	49.89	42.61	0.96	16.74	0.03
Albumin (g/L)	17.94	14.83	0.33	50.58	0.006
Aspartate Transaminase (IU/L)	141.5	200.56	4.61	134.86	0.001
Gamma GT (IU/L)	10.78	51.00	4.28	60.43	0.004
Cholesterol (mmol/L)	2.87	2.14	0.12	18.31	0.02
Triglycerides (mmol/L)	13.53	9.04	0.65	17.88	0.02
Bile acids (μmol/L)	36.87	6.91	1.72	961.11	<0.0001

), but for some of the blood parameters, these changes depended on Diet treatment levels (

Table 5. Significant Diet effect on plasma parameters (36 samples within 12 experimental units).

Plasma Parameters 1	Diets				F Ratio	p-Value
	0% DBSFL 2	10% DBSFL	18% DBSFL	SEM
Total protein (g/L)	49.25a	46.67a	42.83b	0.60	28.68	0.01
Albumin (g/L)	17.92a	16.17ab	15.08b	0.44	10.77	0.04
Triglycerides (mmol/L)	13.05a	12.19ab	8.62b	0.64	13.40	0.03

¹ In each row, means followed by the same letter are not significantly different. ² DBSFL: Dried black soldier fly larvae.

) or Time × Diet interactions (

Table 6. Significant (p < 0.05) ¹ Time × Diet interaction effect on plasma parameters (36 samples within 12 experimental units).

	Diets
Plasma Parameters 1	0% DBSFL 2		10% DBSFL		18% DBSFL		SEM	F Ratio	p-Value
	Week 2 3	Week 17	Week 2	Week 17	Week 2	Week 17
Creatinine (μmol/L)	14.33	15.00	14.83	11.33	12.17	8.67	1.31	12.50	0.04
Potassium (mmol/L)	3.77b	6.15a	3.85b	5.88a	3.90b	5.65a	0.15	1850	<0.0001
Amylase (IU/L)	387.50ab	385.50ab	422.33a	332.00ab	397.50a	277.50b	16.16	10.66	0.04
Creatine Kinase (IU/L)	613.33c	1904.00ab	721.17c	2336.33a	663.67c	1552.83b	64.19	20.52	0.02

¹ In each row, means followed by different letters were significantly different (p < 0.05); ² DBSFL: Dried black soldier fly larvae; ³ At Week 2 of the experiment, hens were 20 weeks old, and hens in all three treatment groups were still on the control diet.

). The lipid content of the 18% DBSFL diet was higher than that of the control diet, but plasma triglycerides were higher in the control hens than in the 18% DBSFL hens (Table 5).

3.7.2. WBC Differential Counts

Most of the WBC differential counts were significantly affected by Time (

Table 7. There was a significant difference in WBC differential counts between 20-week-old hens’ baseline plasma and 35-week-old hens’ blood samples (36 samples within 12 experimental units).

WBC Differential	Age		SEM	F Ratio	p-Value
	20-Week-Old	35-Week-Old
WBC count (×109/L)	26.19	16.18	1.65	10.43	0.046
Hematocrit or packed cell volume (L/L)	0.30	0.26	0.006	56.97	0.004
Basophils (%)	1.24	2.57	0.53	26.97	0.02
Heterophils (%)	27.14	42.96	2.04	54.19	0.005
Lymphocytes (%)	63.17	50.58	1.26	25.54	0.01
Lymphocytes absolute (×109/L)	16.68	7.89	0.77	41.00	0.006
Heterophil:lymphocyte ratios (H:L ratio)	0.54	0.96	0.05	23.74	0.02
Monocytes (%)	7.36	2.55	0.50	24.33	0.01
Monocytes absolute (×109/L)	1.91	0.45	0.13	38.75	0.007

), and, except for monocytes, the effect of the diet and the Diet × Time interaction was not significantly different. The 10% DBSFL hens had significantly higher percent monocytes (5.92 ± 0.18%) than the control hens (3.87 ± 0.26%) (F(2,2.21) = 20.47, p = 0.04). The percent lymphocytes of 18% DBSFL hens were 5.08 ± 0.18%.

3.8. Fecal Microbial Analysis

Results of the fecal culture indicated no Salmonella, Shigella, or Campylobacter contamination in the trailers (or consequently, the treatments). Fecal culture direct exam results are presented in

Table 8. Fecal microbial analysis direct exam results (n = 6).

Diets	Replication (Trailer)	Direct Exam 1
		Gram-Negative Bacilli	Gram Positive Bacilli	Gram Positive Cocci
0% DBSFL 2	1	Moderate	Moderate	Many
	2	Moderate	Scant	Moderate
10% DBSFL	1	Moderate	Moderate	Many
	2	Moderate	Scant	Moderate
18% DBSFL	1	Moderate	-	Moderate
	2	Moderate	Moderate	Many

¹ ‘Many’ indicates that the organism was present in all areas of the culture plates; ‘moderate’ indicates half the plate was covered; and ‘scant’ indicates that the organism was only present in the initial sample inoculum. ² DBSFL: Dried black soldier fly larvae.

.

3.9. Health and Welfare Parameters

There was neither mortality nor any observable signs of morbidity. Plumage conditions were excellent, and there were no wounds (scores 4 out of 4 with no variation) on any of the hens from all diet treatments.

3.10. Post-Mortem Examination of Gastrointestinal Tract

The 18% DBSFL hens showed a higher percent (of the digestive tract) of duodenum weight than control birds (

Table 9. Post-mortem examinations of the gastrointestinal tract in hens on dried black soldier fly larvae (DBSFL) diets.

DI Weights 1	Diet
	0% DBSFL		10% DBSFL		18% DBSFL
	Mean	SEM	Mean	SEM	Mean	SEM
Hen Live weight (kg)	1.93	0.08	1.73	0.04	1.73	0.10
Total DI Weight (g)	109.77	6.05	107.03	2.88	110.11	6.39
DW % BW 2	5.73	0.33	6.19	0.11	6.43	0.43
Proventriculus % BW 2	0.34	0.02	0.34	0.01	0.38	0.02
Proventriculus % DW 3	5.87	0.12	5.52	0.12	5.96	0.08
Gizzard % BW	1.50	0.08	1.66	0.08	1.68	0.15
Gizzard % DW	26.27	0.69	26.90	1.15	26.03	0.90
Duodenum % BW	0.49	0.03	0.55	0.03	0.59	0.03
Duodenum % DW	8.48b	0.22	8.93ab	0.47	9.28a	0.20
Jejunum % BW	0.55	0.06	0.69	0.04	0.69	0.08
Jejunum % DW	9.45	0.57	11.19	0.59	10.59	0.66
Ileum % BW	0.38	0.03	0.49	0.02	0.48	0.06
Ileum % DW	6.65	0.28	7.98	0.30	7.39	0.51
Ceca % BW	0.39	0.03	0.38	0.01	0.37	0.03
Ceca % DW	6.81	0.29	6.19	0.32	5.80	0.20
Colon % BW	0.21	0.02	0.20	0.02	0.17	0.01
Colon % DW	3.69	0.27	3.24	0.30	2.66	0.18
Liver % BW	1.69	0.09	1.69	0.06	1.88	0.10
Liver % DW	29.65	0.38	27.26	0.77	29.47	1.01
Pancreas % BW	0.18	0.01	0.17	0.01	0.18	0.01
Pancreas % DW	3.16	0.14	2.80	0.16	2.83	0.14

¹ DI = Digestive system segments examined; ² BW = hen live weight; ³ DW = DI weight.

).

4. Discussion

Changes in the production efficiency indicators of laying hens are expected when insects are included in their diets as substitute protein and lipid sources [9,10]. In the current study, control hens had a better weight gain, laid more and bigger eggs while consuming less feed, and had a better feed conversion ratio than 18% DBSFL hens. However, the production performance of the 10% DBSFL hens was not significantly different from the control hens in many of the parameters assessed (e.g., HDEP). Previous studies also recommended that the partial substitution of conventional protein sources in poultry diets with insects be feasible [12].

The effects of the use of BSFL meal in poultry feed have been investigated in several studies, while the use of full-fat BSFL has been reported only in a few studies. The fat content of the BSFL meal is extracted using different methods. For example, the larvae are dehydrated and ground, and then the Soxhlet method is used to extract the fat. Alternatively, the grounded larvae are repeatedly washed with a hexane organic solvent [37]. Another method is to use compression [18]. In addition, the inclusion rates and methods (i.e., whether they were used to substitute nutrients in the diet or used as an extra nutrient source alongside a formulated diet), BSFL production and processing methods, poultry production system, and hen age also varied among different studies, which all contribute to the reporting of inconsistent results in different articles. For example, Ruhnke et al. [38] reported that offering full-fat DBSFL ad libitum in an on-range choice feeding trial (from 47 to 62 weeks of age) did not affect performance parameters, but the egg weight was significantly lower in DBSFL hens. Nevertheless, Kawasaki et al. [39] found that the inclusion of full-fat BSFL and black soldier fly pre-pupae (from 24 to 5 weeks of age) did not impact the feed intake or hen egg-laying rates, while the pre-pupae treatment level resulted in the production of heavier eggs than the control and BSFL treatment levels. Moreover, Tahamtani et al. [40] fed four different amounts of live BSFL (0% to ad libitum) to laying hens (from 18 to 30 weeks of age) in addition to their full feed concentrate diet and found that hens (at ad libitum feeding) had higher protein, lipid, and energy intakes and weight gains compared to the other groups; nevertheless, they did not notice any effects on egg production or egg weight.

Similar inconsistencies have also been observed in the results reported from the BSFL meal inclusions in laying hen diets. Maurer et al. [41] used partly BSFL meal to replace soybean meal partially and fully in organic layer diets and fed old, spent White Leghorn layers (64–74 weeks old) for 3 weeks. They concluded that feed efficiency was maintained at a level similar to that of soybean meal-based diets, and the metabolic and health status of the hens were not affected. Marono et al. [42] studied the effects of substituting soybean meal completely with BSFL meal in the diet of 24 to 45-week-old hens. They found that feed intake, lay percentage, egg weight, and FCR declined in the BSFL group compared to the soybean meal-fed group.

The within-treatment live weight of 34-week-old hens in the control and the two DBSFL treatment groups in the current study all exceeded the target weight of 1.83 kg recommended by Novogen [31]. The 18% DBSFL hens consumed more feed compared to the other two groups, but the control hens gained more weight at a faster rate than the 10% and 18% DBSFL hens. As a result, control hens had significantly better FCR compared to the 18% DBSFL hens. Several dietary factors can potentially affect the FCR and performance of the 18% DBSFL hens, as listed below.

• The 18% DBSFL diet was 30% bulkier than the other two diets because of the addition of 10% wheat bran to balance the nutrient provision with the high lipid content of DBSFL in the diet. The bulkiness of a diet impacts the amount of feed consumed (see Waldroup et al. [43]). Leeson et al. [44] reported increased feed intake because of nutrient dilution in bulky feed, which resulted in lower weight gain in layers. Oku et al. [45] demonstrated the inhibitory role of unavailable carbohydrates on the intestinal Ca absorption caused by the loss of Ca-binding protein when large quantities of undigested ingredients transit the gastrointestinal tract. Pelleting bulky diets reduces their bulkiness [46]. This solution can be applied to reduce the volume of full-fat DBSFL diets when using a bulkier ingredient (e.g., wheat bran) is necessary to balance the diets.
• The exoskeleton of the chopped larvae and the tissue sheltered inside may pass through the digestive system without being digested, resulting in the loss of some nutrients. Despines and Axtell [47] reported observations of undigested larval exoskeletons in the excreta of turkey pullets when they were fed darkling beetle larvae, and Zuidhof et al. [48] had similar observations in feeding house fly larvae to turkey pullets. The form of DBSFL (whole, chopped, or ground) therefore affects digestion and absorption of its nutrients [23,24]. Although crude protein levels were similar in the control and 18% BSFL diets, the plasma protein concentration was lower in 18% DBSFL hens, probably because of the lower digestibility and absorption of the nutrients in DBSFL considering the presence of the exoskeleton. Plasma proteins are good indicators of health and contribute to gluconeogenesis, maintenance of colloid osmotic pressure, production of enzymes and immunoglobulins, and transport of minerals and hormones [49]. Kerstetter et al. [50,51] reported the negative impact of low-dietary protein on Ca absorption in the gut. Plasma protein and Ca concentrations declined with increases in DBSFL levels in the present study. The 18% BSFL hens also had thinner egg shells compared to the control eggs [52]. Short-term studies and studies using old or low egg production hens may not be able to detect such Ca deficiency, and this may explain the inconsistency in the results reported from different studies.
• Another potential factor affecting the Ca digestion and absorption of the 18% DBSFL hens could be related to the use of full-fat DBSFL. Using full-fat DBSFL to fully replace soybean meal and the majority of the soybean oil in the diet resulted in a high lipid content in the feed. The negative impact of the higher fatty acid concentrations in feed on the absorption of some nutrients, especially Ca, is well documented [53,54] because Ca can combine with fatty acids in the gastrointestinal tract to form insoluble soaps, which prevent its absorption. The availability of a high level of digestible Ca is crucial for egg production. Although the Ca concentrations were higher than the recommended level in all three experimental diets, the Ca present in the exoskeleton of the whole or chopped DBSFL could have a low bioavailability (see point 4 below). This may result in a faster depletion of the medullary bone Ca a few weeks after the beginning of egg production [55]. The drop in egg production of the 18% DBSFL hens contributed to the FCR calculation. In addition to the drop in egg production, the egg weights of the 18% DBSFL hens were also significantly lighter than the control eggs after the first 4 weeks of the experiment. Further research to determine the digestibility of the Ca in the DBSFL exoskeleton for laying hens may be worthwhile.
• The total percent oil concentrations (crude fat) of the three experimental diets exceeded the fat requirements [30]. However, the plasma triglycerides of the 18% DBSFL hens were lower than those of the control hens, and Marono et al. [42] reported similar results. Avian energy reserves are mostly stored as lipids, specifically triglycerides [56]. Serum triglyceride concentrations are the most sensitive blood metabolites to feed or water deprivation [57]. Hossain and Blair [58] also reported that serum triglyceride concentrations were significantly lower in the birds fed diets containing chitin compared to the control birds. Lower plasma triglycerides in the 18% DBSFL hens may indicate that they were not absorbing the optimally required nutrients.
• Finke [23] reported that the chitin content of the BSFL is 2-3% on an “as-is” basis (i.e., 5 to 7.5% on a dry-matter basis). Marono et al. [42] calculated that full replacement of soybean meal with 17% of BSFL meal (with 5.4% chitin content on a dry-matter basis) resulted in the provision of 1.02 g chitin/day/hen. The presence of chitin in the exoskeleton of DBSFL may influence the digestibility and bioavailability of its nutrients [59,60]. Chitin is insoluble in most solvents and has limited digestibility in poultry diets, and the inclusion of chitin in broiler feed significantly reduced the apparent digestibility of protein and the serum triglyceride concentration [58]. Chitin may also result in a slight overestimation of the crude protein content of the feed because it contains an N-acetyl group in its structure [61]. Biasato et al. [62] documented higher feed intake in diets containing high levels of chitin.
• The results of the present study showed that the 18% DBSFL hens had a heavier duodenum (percent of the digestive tract) than the control hens. Johnson [55] indicated that Ca absorption formation occurs in the duodenum and upper jejunum. Biasato et al. [62] reported that inclusion of the yellow mealworm larvae in broiler chicken diets resulted in shorter villi and deeper crypts in the duodenum. This, in turn, decreased the digestibility and absorption of nutrients and caused poor performance in the birds. The full replacement of the soybean meal with BSFL meal resulted in a higher villi height in the duodenum but a lower villi height in the jejunum and the ileum in laying hens, as well as a change in the enzymatic activities of the brush border membrane [42,63]. The effects of the inclusion of BSFL on the digestibility of nutrients and intestinal enzymatic activities require further investigation.

Furthermore, one of the concerns about the use of DBSFL in poultry feed was related to the possibility of the presence of pathogens in feeds containing DBSFL. The inclusion of DBSFL in the present study did not introduce new pathogens, and the fecal microbial composition was comparable among the treatment levels. Another concern regarding the use of DBSFL has been the potential for higher concentrations of heavy metals in the DBSFL. The 18% DBSFL diet had the lowest concentrations of Cu and Cr compared to the other two diets. In addition, Zn concentrations were similar in all three diets. The As, Cd, and Pb concentrations of all three diets were negligible. The results of the present study revealed that the heavy metal contents of DBSFL were not significant, and none of the heavy metals examined exceeded the reported limits [64]. The heavy metal content of the DBSFL is affected by the sources of feed material provided to the larvae during their production period.

There was no mortality or any observable signs of health disorders in the flocks. Assessing plumage condition and wound scores, all the hens had top scores with no variation. An examination of the WBC counts did not reveal any signs of stress in the hens.

In the current study, hens from all groups had access to the cover crop in the free-range production system. The amount of cover crop consumption was not considered in the calculation of the FCR; however, it was controlled for by providing the same level of access to all groups at all times.

5. Conclusions

In summary, the use of DBSFL in poultry diets had no adverse effects on the health and welfare of the hens. Although 18% DBSFL hens were slower in gaining weight, gained less weight, and laid fewer and smaller eggs than the control hens, they were still gaining weight, maintaining above 90% HDEP, and their eggs were still in the Canadian large egg category.

The use of chopped full-fat DBSFL in the diet increased the bulkiness of the feed. Using BSFL meals may alleviate this problem. The BSFL meal will not be high in lipid content, so it will not be necessary to increase the bulkiness of the feed to balance the diet when substituting for soybean meal in the diets.

In conclusion, the authors recommend partial replacement of soybean meal and oil with full-fat DBSFL in layer diets because, at the partial substitution treatment level of 10% DBSFL, hen health and welfare, feed safety, and production efficiency parameters were comparable to those of the control hens. Further studies are required to determine the most efficient partial substitution rates for the inclusion of DBSFL in poultry diets and to develop new production efficiency criteria to facilitate the use of edible insects and alternative feed ingredients in poultry diets.