Enrichment with Hermetia illucens live larvae in a cage-free system: effects on welfare and egg production and quality in laying hens of two genotypes

Abstract

The effect of the enrichment (10% ad libitum daily feed intake) with live larvae of Hermetia illucens (HI) was evaluated on behavior, egg production and quality from 36 to 40 weeks of age in 1800 laying hens of two genotypes (brown and white) randomly allocated in 8 pens of an aviary system. The larvae enrichment significantly affected space use and behavior, increasing the number of birds on the floor (P < 0.001), the proportion of clean eggs and that of eggs laid in the nest (P = 0.001), while reducing inactivity, floor pecking and the distance hen-operator at the avoidance test in enriched compared to non-enriched pens. As for differences between genotypes, the rate of inactive hens was higher (P = 0.001) in brown compared to white hens, while the rate of hens pecking the floor and showing preening was lower (P < 0.001). The number of piling events was markedly lower in brown compared to white hens (P < 0.001) and brown hens were also more likely to approach a novel object (P = 0.044). Then, brown hens showed higher rates of broken eggs (P = 0.008), fewer eggs laid in nests and more eggs laid on aviary tiers (P < 0.001). Egg quality differences were limited with more blood and meat spots (P < 0.001), lower yolk proportion (P < 0.001) and thicker shells (P < 0.001) in brown compared to white hens. In conclusion, the enrichment with HI live larvae increased species-specific behaviors in both genotypes without substantially affecting egg production and/or quality. On the other hand, the enrichment improved the space use as eggs laid in the nests increased at the expenses of floor eggs, which finally resulted in an improvement of production.

Introduction

In the whole poultry production sector, consumers’ concerns are driving the transition towards systems which can improve animal welfare and production sustainability (EFSA, 2023a, 2023b). As for laying hens, eleven highly-relevant welfare consequences (i.e., negative changes in welfare that result from the effect of one or more factors) have been identified by EFSA (2023a). Among this welfare consequences, (i.e., the animal experiencing stress and/or negative affective states such as pain, fear and/or frustration resulting from a high incidence of aggressive and other types of negative social interactions) is likely to occur in hens kept in all conventional housing systems allowed in the EU (enriched cages, floor systems with single tier and with multitier). Then, “inability to perform exploratory or foraging behaviors” is always observed in enriched cages and can occur to a different extent in floor systems with a single tier or more tiers, depending on the presence/type of litter as well as additional enrichments or possibility of outdoor access. These welfare consequences may produce aggressive interactions among animals, fear responses, injurious pecking, and piling behaviors (EFSA, 2023a).

As a best practice, enrichment can be used in all systems to promote the expression of specie-specific behaviors, reduce group stress, and increase the ability to perform exploratory and foraging behaviors, while improving physical health (Campbell et al., 2019; Son et al., 2022). Feasible enrichments must be easy to implement, economically viable, and without negative impacts on animals, farm management and productive results (Campbell et al., 2019; Xu et al., 2022). Different types of enrichment can be used, structural and/or feeding ones, with different roles, i.e. favoring physical development of animals (Hartcher and Jones, 2017; Lourenço da Silva et al., 2021), improving space use and animal distribution in cage-free systems (Campbell, 2023), promoting hens’ positive behaviours (Bellezza Oddon et al., 2024). Among the different enrichments, live insects have a high potential for promoting poultry natural feeding behavior (Schiavone and Castillo, 2024) besides production sustainability (Dörper et al., 2021). Among insect species eligible for such enrichment, live larvae of black soldier fly (Hermetia illucens, thereafter HI) have a great potential in view of sustainability, due to high bioconversion ability and adaptability of this species to a wide range of organic substrates, besides being a valuable feed as their crude protein content can reach 13% with lipid content averaging at 10% (Cattaneo et al., 2025). In fact, previous studies using local breeds kept under cage-free systems with outdoor access have found that enrichment with live HI larvae increased comfort (such as preening and allo preening) (Bellezza Oddon et al., 2024) and foraging behaviors (Tahamantani et al., 2025) and improved feather conditions (Bellezza Oddon et al., 2024; Stars et al., 2020) of laying hens. Larvae provision can also affect gut health of laying hens (Bellezza Oddon et al., 2024; Huang et al., 2024), where such enrichment could definitively promote welfare also in commercial genotypes of laying hens kept in cage-free systems without outdoor access (Sokołowicz et al., 2020). In fact, even within commercial genotypes, the genetic background has a substantial impact on the behavioral responses, emotional reactivity, environmental adaptability, space use and finally welfare and health conditions of farmed laying hens (Sokołowicz et al., 2020; Ciarelli et al., 2023; Gandarillas et al., 2025). which is pivotal under cage-free systems and/or free-range conditions where genotype, management, and housing systems strictly interact.

Then, using insects as an enrichment in laying hens can also have consequences on production, due to their nutrient supply, besides egg quality nutritional and sensorial traits because of the specific contents in fatty acids and carotenoids (Marín et al., 2024; Papin et al., 2025).

Thus, the present study aimed at evaluating the effects of a feeding enrichment based on the supplementation of HI larvae offered at 10% of daily ingestion on the behavior and space use of laying hens, besides product quality in two commercial genotypes, brown-feathered and white-feathered laying hens, reared in a multi-tiered aviary system from 36 to 40 weeks of age. Fecal microbiota was also evaluated as a welfare indicator at the gut level.

Material and methods

Housing and management of laying hens

This study run in the facilities of the Experimental Farm of the University of Padova, Italy, following approval by the Ethical Committee for Animal Experimentation (Organismo per la Protezione del Benessere Animale, OPBA) of the same University (Project 28/2020, prot. 204398, approved on 06/05/2020). All animals were handled according to the principles stated by the EU Directive 2010/63/EU regarding the protection of animals used for scientific purposes. The researchers involved in animal handling were either animal specialists (PhD or MSc in Animal Sciences) and/or veterinary practitioners.

The experimental aviary (2.50 m wide × 19.52 m long × 2.83 m high) consisted of 3 tiers and a litter area on the ground. Two corridors were available for the hens adjacent to the two sides of the aviary, so the floor space was 6.01 m wide × 19.52 m long. The aviary was divided into 8 pens (2.50 m wide × 2.44 m long × 2.83 m high), each one hosting 225 animals at the housing time for a density of 9 hens/m² of available surface (Fig. 1). Each pen was equipped with 4 collective nests. A programmable logic controller (Officine Facco & C. Spa, Campo San Martino, Padova, Italy) managed the automatic systems for feeding, drinking, lighting, and ventilation inside the system.

Fig. 1.

Design of a single module of the multi-aviary system.

During the experimental period, the barn temperature was maintained at 22.0 ± 0.5°C with relative humidity at 76.8 ± 4.8%; hens were exposed to a 16-h light/8-h dark photoperiod with dusk and down phases.

The commercial diet was mainly based on corn and soybean meal and formulated to meet the nutritional requirements of laying hens (Table 1, Table 2) (Novogen, 2025a, 2025b).

Table 1.

Proximate composition (% as fed) and mineral content (mg/kg) of the diet and the live larvae of Hermetia illucens (HI) used as feeding enrichment.

	Diet1	Live HI larvae
Proximate composition
Dry matter (%)	90.4	30.6
Crude protein (%)	17.0	11.3
Ether extract (%)	6.41	9.38
Crude fiber (%)	4.74	-
Starch (%)	25.8	-
Chitin (%)	-	1.95
Ash (%)	15.5	2.43
Mineral composition (mg/kg)
Ca	42932	5077
Fe	397	73.0
K	9668	3463
Mg	2238	994
Mn	193	35.5
Na	3232	376
P	7858	2591
S	2749	956
Zn	113	35.6

¹Ingredients: Corn, Soybean solv. extr. meal, Calcium carbonate, Full-fat soybean, Corn gluten meal, Monocalcium phosphate, Oat bran, Wheat meal, Soybean oil, Chabasite zeolite, Sodium chloride, Sodium bicarbonate, Vitamin-mineral premix (provided per kg of diet: Vitamin A 8 000 IU, Vitamin D3 2 500 IU, Vitamin E 25 mg, Choline 350 mg, Fe 35 mg, Cu 8 mg, Mn 105 mg, Zn 42 mg, I 0.50 mg, Se 0.30 mg, DL-methionine 1470 mg, 6-phytase 600 FYT, Endo-1,4-beta-glucanase 250 TGU, Endo-1,4-beta-xylanase 560 TXU).

Table 2.

Fatty acid (FA) profile (% of total FA) of the diet fed to laying hens and live Hermetia illucens larvae used as a feeding enrichment.

Fatty acid (% total FA)	Diet	Live HI larvae
C12:0	0.10	41.3
C14:0	0.15	7.60
C16:0	12.1	11.0
C18:0	3.66	2.04
Other SFA	0.76	1.30
C16:1 n7	0.12	3.03
C18:1 n9	22.6	13.7
C18:1 n7	1.22	0.35
Other MUFA	0.26	0.85
C18:3 n3	5.99	1.72
C18:2 n6	52.8	16.0
Other PUFA	0.25	1.10
Total SFA	16.8	63.3
Total MUFA	24.2	17.9
Total PUFA	59.0	18.8
PUFA n3	6.16	1.80
PUFA n6	52.9	17.7
PUFA n6/ PUFA n3	8.59	9.68

SFA; saturated FA; MUFA: monounsaturated FA; PUFA: Polyunsaturated FA; LA: linoleic acid; GLA: gamma-linolenic acid; ALA: alpha-linolenic acid; DPA: docosapentaenoic acid; DHA: docosahexaenoic acid.

Animals

At 17 weeks of age (114 days), 1800 hens, half of which brown-feathered (Novogen Brown) (initial live weight: 1304 g ± 129) and half White-feathered (Novogen White) (live weight: 1172 g ± 90.8) (Novogen S.A.S., Rue des Compagnons, Secteur du Vau Ballier, Pledran, France) were delivered to the experimental farm by an authorized truck from a commercial pullet house using a floor system with platforms for pullet training. Hens had been previously vaccinated against Marek disease, Newcastle disease, bronchitis, coccidiosis, laryngotracheitis, salmonellosis, mycoplasmosis, avian typhoid, colibacillosis, and encephalomyelitis.

Upon their arrival, pullets were housed in the eight pens of the experimental aviary, where pens with brown-feathered hens were alternated with those with white-feathered hens. The animals were initially housed under a 12-hours light/12-hours dark cycle, which was increased by half an hour of light per week until reaching 16 hours of light and 8 hours of darkness at 25 weeks of age (169 days). Egg production started at 20 and 22 weeks of age and oviposition rate per hen exceeded 80% at 23 and 25 weeks of age in white-feathered and brown-feathered hens, respectively.

Experimental setup

For the specific purposes of the present study, a 5-week trial was implemented when hens were 36 weeks of age. At this age, the number of hens per pen ranged from 215 to 223 because of mortality occurred since housing. During this trial, out of the eight available pens, four pens of the aviary (two with brown-feathered hens and two with white feathered hens) received only the commercial diet (Control group) and the remaining four (two with brown-feathered hens and two with white-feathered hens) received a supplementation with live HI larvae (HI group) (10% of the average daily feed intake of the whole group) once per day (around 10:00 am). The live HI larvae were manually distributed in two additional circular feeders (diameter: 37 cm) per pen placed on the floor area on the four enriched pens.

As for HI larvae, the amount requested for supplementation was produced by BEF Biosystem (Turin, Italy) and delivered to the experimental farm every 15 days. The larvae were transported over the course of one day and shipped together with their substrate. Upon arrival, the live larvae were stored in a refrigerated room at 14°C under dark conditions to reduce their metabolic activity, halt development, and ensure the availability of live larvae for the subsequent 14 days. At each delivery, 200 g of larvae were collected and stored at −20°C for subsequent laboratory analyses. Prior to administration, the larvae underwent a revitalization phase, during which they were maintained in a conditioned room at 28°C for 10 minutes. This process reactivated their metabolism and motility, which maintained their attractiveness as environmental enrichment for the hens (Bellezza Oddon et al., 2021). The level of inclusion was stated based on available literature (Tahamtani et al., 2021; Dörper et al., 2024).

Spontaneous behaviors

The real-time video recording system used a total of 48 cameras (Infrared mini-dome bullet 4 mp; resolution 1080 p) (HAC—HDW1220MP; Zhejiang Dahua Technology Co. Ltd., Hangzhou, China) and two full HD video-recorders (NVR2116HS-4KS2; Zhejiang Dahua Technology). The cameras were located to record the hens on the ground (one camera per pen hanged at about 3 m of height) and the hens on the other levels of the aviary (first, second and third tiers). The system was set up to get and store 24-h videos once per week using all the cameras for the 5 weeks of trial. Then, videos recording hens on the ground were scanned by two trained people for a maximum of 15 seconds every 30 minutes from 5:00 am to 7:30 pm, with the exception of the observation at time of the HI larvae delivery, where videos were scanned 15 min later (i.e., at 10:15 am). The number of hens on the litter and the number of hens performing different behaviors at the beginning of the observation were scored. In detail, the following mutually exclusive behaviors (Nielsen et al., 2003; Thuy Diep et al., 2018; Trocino et al., 2020) were scored and then expressed as rate of hens performing the behaviour out of the total number of hens observed on the floor per observation time:

• −pecking at the feeder: pecking at the feeder;
• −pecking the floor: pecking on the ground;
• −dust-bathing: hen forced the sand or other materials into the plumage by squatting on the ground and making appropriate movements with the body, wings, and legs;
• −inactive: standing or sitting with no other activity;
• −preening: hen directs its beak to its own plumage of several body parts (thorax, abdomen, shoulder, interior and exterior wings, rumps, back, and cloaca) and carries on pecking, nibbling, combing or rotating movements, once or repeatedly;
• −moving: standing or walking (running), with no other pecking or scratching activity;
• −piling: more than three mostly immobile (maximal movement duration < 5 second) hens standing in the closest possible proximity (overlapping of body outlines) with most hens facing in the same direction;
• −aggression: pecking or fighting;
• −others: unable to identify the state of the hen in any of the previous ones.

Inter-observer agreement, assessed on a subset of observations independently scored by both observers using the Intraclass Correlation Coefficient (ICC > 0.6 for all behaviors).

Novel object test

The novel object test (Welfare Quality Project, 2009) was performed in all pens on the last week of trial, when the hens were 40 weeks old, following the order of the pens, i.e. first the four pens receiving the enrichment with live larvae and then the four pens not receiving the enrichment, with pens containing brown hens alternating with those containing the white hens. The novel object, consisting of a 50 cm-long stick with colored bands and a diameter of approximately 2.5 cm, was placed centrally and with slow movements within the littered area accessible to the animals from the outside corridor of the aviary without the operator entering the pen. The object remained within the pens for about 3 minutes, and then it was cautiously retrieved by the operator from the outside corridor of the pen. This procedure was repeated for each pen. During the test, the behavior of laying hens on the floor was video recorded. Then, videos were scored to check the number of hens located within one body length (≈30 cm) from the novel object every 10 seconds during the first two minutes, starting from the exact moment of the novel object placement in the pen.

Avoidance distance test

The avoidance distance test (Welfare Quality Project, 2009) was performed in all pens on the last week of the trial, two days after the novel object test following the same pen order of the novel object test. The same operator entered each pen walking slowly through the center of the littered area, maintaining a calm and steady pace. He held his hand in a fixed position in front of the abdomen, directly towards the animals starting from those on the first tiers and moving to those on the second tiers of the aviary (if the number of hens on the first tier was not sufficient) considering one hen per time for a maximum of 5 hens per pen. He randomly selected one hen per time, which was observed sitting on the edge of the aviary structure, and advanced towards the hen at a consistent pace of one step per second, keeping his gaze focused on the hen toes. As soon as the hen moved away or turned (i.e., shifted both feet sideways or backwards), the distance was measured in centimeters between the operator`s hand and the hen`s original foot position.

Production

At the beginning and at the end of the experimental period, 50 hens per pen were randomly taken to measure their live weight using an electronic balance (precision 1 g; Wunder, Sa.Bi. srl, Milan, Italy).

During the whole experimental period, the numbers of total eggs, dirty and broken eggs, eggs in the nest, and eggs on the floor were counted twice per week in each pen. Number of edible eggs was calculated as number of broken and dirty eggs subtracted from number of total eggs. The egg production over the week was calculated at the pen level as:

$$\text{Oviposition}\text{rate}\left(\%\text{present}\text{hen}\right)=\text{total}\text{eggs}\text{at}\text{the}\text{two}\text{weekly}\text{recordings}/\text{total}\text{number}\text{of}\text{present}\text{hens}\text{at}\text{the}\text{two}\text{recordings}\times 100.$$

Egg quality: sampling and analyses

At the end of the experimental period, i.e., after five weeks of HI supplementation, 50 eggs per pen were randomly sampled from those laid on the nest belt, for a total of 400 eggs to be used for measuring their physical and rheological quality and immediately processed in the departmental laboratories for the chemical analysis. Out of the 400 eggs, 240 eggs were used for physical and rheological analysis, 80 were used for shell strength analyses and the remaining 80 for sensory analyses.

The first 240 eggs (30 per pen) were individually weighed and candled to determine the occurrence of any cracks (Simons et al., 2017). The egg equator diameter and height were measured using a digital caliper (IP54, SHAHE Digital Calliper, Wenzhou, Zhejiang, China) and used to calculate the egg shape index (SI) as: SI (%) = egg equator diameter/egg height × 100 (Sirri et al., 2018a). Color measurement of the shell was performed based on the CIE L*a*b* color system using a Minolta Chroma Meter (CM-508, Minolta Corp., Ramsey, NJ, USA). Then, eggs were opened to separate the yolk, albumen, and shell. Egg shells were oven-dried for 24 h at 50°C and weighed (Lordelo et al., 2017). After drying, the thickness of the eggshells (with membrane) was measured at 3 points (the air cell, the sharp end, the equator) using a digital caliper (IP54, SHAHE Digital Calliper). Yolk weight was measured, whereas albumen weight was calculated as the difference between the weight of the whole egg and those of yolk and dry shell. The occurrence of blood and meat spots in the yolk and albumen was visually detected. Yolk color was determined in the CIE L* a* b* color system using a Minolta Chroma Meter CM-508. Yolk and albumen pH were determined using a pH meter (Basic 20, Crison Instruments SpA, Carpi, Modena, Italy), equipped with a specific electrode (cat. 5232, Crison Instruments SpA).

For each pen, out of the 30 eggs per pen separated in yolk, albumen and shell, 10 eggs per pen were used for chemical analyses obtaining 5 pools per pen of 2 yolks each and 5 pools of 2 albumens each (40 pools in total; 10 per experimental group) and freeze-dried to be later analyzed for proximate composition, fatty acid profile, and mineral content).

Shell fracture

Shell fracture test was performed on 10 eggs per pen (80 eggs in total; 20 eggs per experimental group) at the two ends of the eggs, using a texture profile analyzer (Stable Micro System Ltd., Goldaming, UK) equipped with a needle at the constant cross-head speed of 1 mm/sec using a 250 N load cell.

Sensory analysis of egg

Sensory analysis used 10 eggs per pen (80 eggs in total; 20 per experimental group) in a triangle test with 67 panelists in accordance with ISO 4120:2004 guidelines to assess whether participants could distinguish between two samples. Sample triplets consisted of either two eggs from the Control group and one egg from laying hens fed HI larvae, or one egg from the Control group and two eggs from laying hens fed live HI larvae. All triplets were within the same genotype. The presentation order of the samples was randomized across participants. The volunteers, students and staff members of DAFNAE, consumed three eggs coded with three-digit random numbers.

To this purpose, eggs were placed in water at room temperature in a water bath; the water was raised to the boiling point, and the eggs were kept in the boiling water for 8 min. Then, the eggs were removed from the water bath and cooled to room temperature, shelled, and cut into quarters (lengthwise). Samples were served on a 15-cm white paper board plate identified with a 3-digit blind code and each participant was asked to recognize the different eggs. Mineral water and unsalted crackers were provided to panelists to clean the palate between the samples.

Chemical analyses of diets, larvae and eggs

The commercial diet, the HI freeze-dried larvae and the pools of freeze-dried yolk and albumen were analyzed for dry matter content (950.46), ashes (920.153), crude protein (981.10), ether extract (2003.05), crude fiber (987.10) (only the commercial diet) and minerals (2014.004) according to AOAC (2000) methods. The chitin content of HI larvae was measured based on Stelmock et al. (1985). In detail, freeze-dried milled larvae were decalcified in 20 ml of ethylenediaminetetraacetic acid (EDTA) during 15-h incubation, followed by centrifugation for 5 min at 3000 rpm and then decanted. The residues from every sample were transferred to 600 ml Berzelius beakers and refluxed at 90°C in 200 ml 2 N sodium hydroxide for deproteination. The hot beakers were vacuum filtered through glass crucibles and washed three times with distilled water. The residues were dried to a constant weight in a 60°C forced-draft oven and ashed in a muffle oven at 550°C for 2 h. The content of chitin was calculated as: (oven dry weight of residue − residual ash weight) / original dry sample weight × 100.

As for fatty acid (FA) profile measured in diet, HI larvae and freeze dried yolks, fat was extracted from samples by accelerated solvent extraction (ASE; Dionex, Sunnyvale, CA, USA, Application Note 334) using three extraction cycles with chloroform:methanol (2:1 vol) as a solvent at 80°C and a 1-min heating phase and 40‑sec extraction phase (as modified from Folch et al., 1957). The solvent was evaporated under a N₂ stream (Genevac EZ-2, SP Industries, Warminster, PA, USA) at 60°C; the residual samples (extracted lipids in vials) were weighed before adding 4 mL of 1% H₂SO₄ in methanol (Christie, 1998) and held at 50°C overnight. Then, hexane (1 mL per 20 mg extracted fat) and 4 mL of NaSO₄ (0.47% in H₂O) were added and vigorously agitated to transfer the methylated fatty acids in the organic phase. The organic phase was collected after centrifugation and analyzed by GC-FID with an Agilent 7820A Gas Chromatograph (Agilent Technologies, Santa Clara, CA, USA). In details, 1 μL was injected with a split ratio of 65:1. A Supelco OMEGAWAX-TM 250 (Sigma-Aldrich, St. Louis, MO, USA) (30 m × 0.25 mm internal diameter, 0.25 μm film thickness) was used with hydrogen as the carrier at 1.4 mL/min. The oven temperature was set at 50°C, held for 2 min, raised to 220°C at the rate of 4°C/min, and then held for 23 min. Both the injector and the detector temperatures were set at 250°C. The individual FA were identified by comparing the retention time of the standard FA methyl esters mixture (Supelco 37–component FAME Mix, 47,885–U). Individual FA methyl esters were expressed as the percentage of the total area of eluted FA methyl esters.

Mineral content was determined on diets, HI larvae and freeze-dried pools of both yolk and albumen. Briefly, for each sample, 0.30–0.35 g of sample was weighed and placed in a TFM vessel with 2 mL of 30% hydrogen peroxide and 7 mL of concentrated (65%) nitric acid, both of Suprapur quality (Merck Chemicals GmbH, Darmstadt, Germany). These samples were subjected to microwave digestion (Ethos 1600, Milestone S.r.l., Sorisole, BG, Italy) as follows: Step 1, 25–200°C in 15 min at 1200 W with P max 100 bar; Step 2, 200°C for 18 min at 1200 W with P max 100 bar; and Step 3, 200–35°C in 15 min. After cooling to room temperature, the dissolved sample was diluted with ultrapure water (resistivity 18.2 M Ω cm at 25°C) to a final volume of 25 mL. The mineral contents were determined with a Spectro Arcos EOP inductively coupled plasma–optical emission spectrometry (ICP-OES) (Spectro Analytical Instruments GmbH, Kleve, Germany). Calibration standards were prepared using multi- and single-element standard solutions (Inorganic Ventures Inc., Christiansburg, VA, USA) in 30% Suprapur nitric acid (Merck Chemicals GmbH) to obtain similar matrices to the samples.

Sampling, 16S rRNA gene sequencing and bioinformatic analysis for microbiota

Three samples of fresh feces laid on the full surfaces of the aviary per pen (i.e., 24 samples) were collected by a single operator and put in 2-ml Eppendorf and stored at −80°C until analyses. The DNA was extracted from 250 mg of each sample with a Qiagen DNA stool mini kit (Qiagen, Hilden, Germany), following the manufacturer’s instructions. The library preparation for the amplification of V3-V4 regions involved the use of 341F 5 CCTAYGGGRBGCASCAG-3′ and 806R 5- GGACTACNNGGGTATCTAAT-3′ primers pairs, according to methodology analytically described in detail by Klindworth et al. (2013). The amplified 16S rRNA amplicons from each sample were paired-end sequenced on the Illumina NovaSeq 6000 platform according to the 16S Metagenomic Sequencing Library Preparation protocol by Novogene UK company (Novogene, Milton, Cambridge, UK).

Raw reads were processed to remove primer pairs on both ends using the Cutadapt (Martin, 2011) and analyzed using QIIME2 v2021.4 (Bolyen et al., 2019). The ‘qiime dada2 plugin’ was used to denoise and obtain high-quality reads to cluster into amplicon variants (ASVs) with the utilization of SILVA SSU v138.1 database as reference for the taxonomic assignation of ASVs (Quast et al., 2012). ASVs occurring less than 10 times were eliminated.

Statistical analysis

Data were first tested for normality using the Shapiro-Wilk test and for homoscedasticity using Levene’s test. Residuals were modelled assuming a normal distribution. Then, number of animals on the litter, rate of animals performing different behaviors, man-animal distance in the avoidance distance test, and number of animals approaching the object during the novel object test were submitted to analysis of variance (ANOVA) with live larvae enrichment, genotype, week of recording (not for the novel object and the avoidance distance tests) and their interaction as the main factors of variability and the pen as a random effect, using the PROC GLIMMIX procedure of SAS (SAS Institute, 2013). The same approach was used for data related to oviposition rate and rate of eggs lais in the different positions of the aviary. Then, data related to egg quality traits were submitted to ANOVA with live larvae enrichment, genotype, and their interaction as the main factors of variability and the pen as a random effect, using the PROC GLIMMIX procedure of SAS (SAS Institute, 2013). The same main factors were used for the analysis of data related to the chemical composition of eggs using the PROC GLM procedure of SAS (SAS Institute, 2013) and for testing differences in the occurrence of shell pre-cracks, blood spots, and meat spots by the χ2 test of SAS (SAS Institute, 2013). Responses of the triangle test in the sensory analysis were analyzed using a one-tailed binomial test of Fizz Calculations software version 2.50 (Biosystems, Couternon, France), with the null hypothesis assuming a probability of correct selection by chance (P = 1/3). The test assessed whether the proportion of correct responses was significantly greater than expected by chance. Differences between the means with P ≤ 0.05 were considered statistically significant.

As for bioinformatic analysis of microbiota, the microeco (v1.44.0) (Liu et al., 2021) package in R software was used after a TSS (total sum scaling) normalization of the data. The alpha diversity (microbial richness and diversity) was estimated using the Shannon index and differences according to the experimental groups were tested by ANOVA and the Duncan post hoc test. Beta diversity among samples was tested using PCoA (Principal Coordinate Analysis) based on the Bray-Curtis distances. Then, Permutational MANOVA (PERMANOVA) was applied to evaluate the effect of the treatment. Microbial differences and taxa (microbial biomarkers) between treatments were compared by the linear discriminant analysis (LDA) and the effect size (LEfSe) analysis. The Kruskal–Wallis test and the unpaired Wilcoxon test were applied, using an LDA score (log10) > 4, to detect potential biomarkers.

Results

Behavioral recordings

While no significant effect of the interaction between the enrichment and the genotype was recorded (Table 3), the enrichment with HI live larvae significantly increased the number of hens that were found on the floor compared to the non-enriched (control) pens (24.1 vs. 22.9 hens; P < 0.001) (Table 3). As for behaviors, in enriched compared to not enriched pens, on average of all observations, hens were obviously found pecking larvae (6.83% of hens observed on the floor; P < 0.001) which was associated with decreased rates of hens pecking the floor (52.8% vs. 55.6%; P = 0.015), inactive (7.47% vs. 8.40%; P = 0.001), and moving (16.8% vs. 19.0%; P = 0.024). On the other hand, the HI enrichment did not affect the rate of hens piling and showing aggression (on average 2.12% and 0.35%, respectively), dustbathing and preening (on average 1.81% and 12.15% respectively) recorded on the floor (Table 3). As for differences between genotypes, the number of hens observed on the floor was lower in pens with brown hens compared to those with white hens (23.1 vs. 23.8; P < 0.001) (Table 3). Then, the rate of hens pecking the floor (56.6% vs. 51.9%; P < 0.001) and inactive (10.2% vs. 5.68%; P = 0.001) was higher in brown compared to white hens, which corresponded to a lower rate of brown hens preening (10.1% vs. 14.2%; P < 0.001) (Table 3). No significant difference in the rate of aggression was observed between genotypes, whereas the rate of hens piling was much lower in brown hens compared to white ones (0.64% vs. 3.60%; P < 0.001).

Table 3.

Number of animals observed on the floor, rate of animals performing visible behaviors in laying hens receiving or not live Hermetia illucens (HI) larvae (10% daily feed intake) as feeding enrichment of two genotypes (brown and white) from 36 to 40 weeks of age.

	Enrichment (E)		Genotype (G)		Week (W)					P value
	Control	HI	Brown	White	36	37	38	39	40	E	G	W	E × G	E × W	G × W	E × G × W	RMSE
Observations, n	620	620	620	620	248	248	248	248	248
Hens on floor (n)	22.9	24.1	23.1	23.8	24.2b	24.8b	23.8ab	21.7a	22.9a	<0.001	<0.001	0.047	0.169	0.190	0.664	0.793	12.6
Pecking larvae (%)	0.00	6.83	3.53	3.30	5.68b	3.56ab	2.53a	2.65a	2.65a	<0.001	0.586	<0.001	0.633	<0.001	0.742	0.742	8.59
Pecking floor (%)	55.6	52.8	56.6	51.9	58.2b	53.7a	50.9a	53.8a	54.5ab	0.015	<0.001	0.003	0.603	0.010	0.002	0.179	21.5
Dustbathing (%)	2.06	1.56	1.99	1.63	1.93	1.87	1.87	2.07	1.31	0.943	0.481	0.379	0.855	0.439	0.928	0.888	4.47
Inactive (%)	8.40	7.47	10.2	5.68	4.56a	6.11a	9.75b	10.0b	9.23b	0.001	<0.001	<0.001	0.550	0.037	0.006	0.440	9.70
Preening (%)	12.7	11.6	10.1	14.2	10.2a	13.9b	13.3b	11.4ab	12.0ab	0.161	<0.001	0.005	0.962	0.303	0.972	0.534	11.7
Moving (%)	19.0	16.8	16.5	19.3	16.7	18.8	18.9	18.4	16.8	0.024	0.126	0.459	0.515	0.279	0.228	0.629	17.6
Piling (%)	1.88	2.35	0.64	3.60	1.96	1.58	2.53	1.40	3.12	0.329	<0.001	0.209	0.455	0.558	0.485	0.627	9.19
Aggression (%)	0.24	0.45	0.40	0.30	0.47	0.49	0.22	0.28	0.28	0.804	0.554	0.569	0.509	0.725	0.568	0.298	2.27
Others (%)	0.07	0.11	0.06	0.12	0.25	0.07	0.07	0.00	0.06	0.392	0.229	0.030	0.192	0.287	0.849	0.353	0.882

RSME: root mean square error.

Control: not enriched with live Hermetia illucens larvae; HI: receiving live Hermetia illucens larvae (10% daily feed intake) as a feeding enrichment.

As for changes during the five weeks of the trial, the main differences were recorded between the first one or two weeks and the following ones (Table 3). In details, the number of hens found on the floor decreased with the week of trial (24.5 on average of the first two weeks vs. 22.3 on average of the last two weeks of the trial; P = 0.047); then, the highest rates of hens pecking the larvae (5.68% of hens observed on the floor in the first week vs. 2.61% on average in the last three weeks; P < 0.001), and pecking the floor (58.2% of hens observed on the floor in the first week vs. 53.2% on average in the following four weeks; P = 0.003) were recorded at the beginning of the trial, whereas the rate of inactive animals increased as the trial went on (5.33% on average of the first two weeks vs. 9.66% on average of the last three weeks of the trial; P < 0.001). As for preening, the rate of hens showing this behavior was lower at the beginning (36 weeks of age) and at the end of the trial (39 and 40 weeks of age) compared to the other periods (P = 0.003) (Table 3).

As for the rates of hens pecking the floor and inactive, significant interactions were observed between the week of trial and the enrichment (Fig. 2) and the genotype (Fig. 3). In detail, the highest rate of hens pecking the floor was recorded on the first week of the trial (36 week of age) in not-enriched pens compared to all other observations (P = 0.010) (Fig. 2a). Then, as for the rate of inactive hens, no significant difference was observed between the enriched and not-enriched pens during the first week of trial; then, a higher rate of inactive hens was recorded in not-enriched compared to enriched pens on the second and fourth week of trial, while an opposite trend was recorded on the last week of trial (P = 0.037) (Fig. 2b).

Fig. 2.

Rate of laying hens (% hens observed on the floor) pecking the floor (a) and inactive (b) (average of brown and white-feathered hens) receiving or not live Hermetia illucens (HI) larvae (10% daily feed intake) as a feeding enrichment (significant interaction Week of age × Enrichment with HI live larvae).

Fig. 3.

Rate of laying hens (% hens observed on the floor) pecking the floor (a) and inactive (b) in brown and white-feathered hens (average of hens receiving or not live Hermetia illucens larvae, 10% daily feed intake) (significant interaction Week of age × Genotype).

As for interactions between the week of age and the genotype (P = 0.002), the rate of hens pecking the floor was higher in pens with brown hens compared to those with white hens only during the first two weeks of trial (Fig. 3a). Then, the rate of inactive hens did not differ between genotypes in the first two weeks whereas it became significantly higher in brown vs. white hens during the last three weeks of the trial (P = 0.006) (Fig. 3b).

Reactivity tests

At the avoidance distance test, the distance between the operator and the hens was lower in enriched compared to not-enriched control pens (23.3 cm vs. 55.6 cm; P < 0.001) and in pens with brown hens compared to those with white hens (12.6 cm vs. 66.5 cm; P < 0.001) (data not reported in tables). Nevertheless, a significant interaction between enrichment and genotype was recorded (Fig. 4). In detail, the highest distance hen-operator was measured in not-enriched pens with white hens compared to the pens of the other experimental groups (98.6 cm vs. 12.9 cm in not-enriched control pens with brown hens, 34.4 cm in enriched pens with white hens, and 12.3 cm in enriched pens with brown hens; P < 0.001).

Fig. 4.

Distance (cm) between the operator and the hens at the avoidance distance test in laying hens after five weeks of trial: brown-feathered Control hens not receiving the enrichment (Brown-Control); white-feathered Control hens not receiving the enrichment (White-Control); brown-feathered hens enriched with live Hermetia illucens larvae (Brown-HI; 10% daily feed intake as feeding HI from 36 to 40 weeks of age); white-feathered hens enriched with live Hermetia illucens larvae (White-HI) (significant interaction Genotype × Enrichment).

At the novel object test, no difference in the number of hens that approached the object was recorded between control not-enriched and enriched pens (1.75 hens, on average) (Fig. 5a) whereas more brown hens approached the novel object than white hens did (2.06 vs. 1.44; P = 0.044) (Fig. 5b).

Fig. 5.

Number of hens approaching the object during the novel object test performed in laying hens after five weeks of trial: a) not-enriched Control hens vs. hens enriched with live Hermetia illucens larvae (HI; 10% daily feed intake as feeding enrichment from 36 to 40 weeks of age); b) brown-feathered hens vs. white-feathered hens.

Egg production

The enrichment with live HI larvae did not affect hen live weight at the end of the experimental period (1.77 kg vs. 1.79 kg in enriched vs. not enriched hens; RMSE: 0.13 kg; P = 0.197) (data not reported in table), the oviposition rate referred to all laid eggs (95.7% vs. 94.0% of present hens in enriched vs. not enriched hens; P = 0.100), while increased the rate of edible eggs (93.0% vs. 89.1% in HI hens compared to control hens; P = 0.001) due to decreased rate of dirty eggs (2.45% vs. 4.77%; P = 0.013) (Table 4). In fact, out of all laid eggs, the proportion of eggs laid in the nests was higher (97.7% vs. 95.1%; P = 0.001) while the proportion of floor eggs was lower (0.59% vs. 3.23%; P = 0.002) in enriched pens compared to not-enriched control pens (Table 4).

Table 4.

Oviposition rate (% present hens) of total, edible, dirty, and broken eggs and position of eggs laid in laying hens receiving or not live Hermetia illucens (HI) larvae (10% daily feed intake) as a feeding enrichment of two genotypes (brown and white) from 36 to 40 weeks of age.

	Enrichment (E)		Genotype (G)		Week (W)					P value							RMSE
	Control	HI	Brown	White	36	37	38	39	40	E	G	W	E × G	E × W	G × W	E × G × W
Oviposition rate
All eggs (%)	94.0	95.7	94.5	95.2	95.7	94.7	94.7	94.6	94.4	0.100	0.487	0.930	0.255	0.101	0.113	0.310	3.33
Edible eggs (%)	89.1	93.0	90.4	91.7	92.2	91.4	90.8	90.7	90.0	0.001	0.225	0.749	0.275	0.336	0.123	0.485	3.31
Dirty eggs (%)	4.77	2.45	3.82	3.40	3.20	3.03	3.88	3.80	4.15	0.013	0.627	0.904	0.965	0.451	0.676	0.981	2.80
Broken eggs (%)	0.16	0.29	0.32	0.13	0.34	0.29	0.06	0.17	0.26	0.080	0.008	0.100	0.849	0.030	0.493	0.558	0.20
Eggs positions
Nest (%)	95.1	97.7	94.4	98.4	97.5	96.0	96.4	96.1	96.1	0.001	<0.001	0.619	0.964	0.891	0.926	0.920	2.99
Tier (%)	1.68	1.66	3.12	0.22	1.14	1.86	1.71	1.86	1.77	0.972	<0.001	0.638	0.892	0.872	0.413	0.923	1.96
Floor (%)	3.23	0.59	2.47	1.36	1.32	2.13	1.94	2.04	2.13	0.002	0.156	0.954	0.986	0.897	0.911	0.935	3.18

RSME: root mean square error.

Control: not enriched with live Hermetia illucens larvae; HI: receiving live Hermetia illucens larvae (10% daily feed intake) as a feeding enrichment.

As for the effect of the genotype, live weight of brown hens was higher compared to white hens (1.90 kg vs. 1.66 kg; RMSE: 0.14 kg; P < 0.001) (data not reported in table) without differences between genotypes as for the oviposition rate of all laid eggs (94.8% of present hens on average; P = 0.487). Then the oviposition rate of broken eggs was higher in brown hens compared to white hens (0.32% vs. 0.13%; P = 0.008) (Table 4). Indeed, the proportion of eggs laid in the nests was lower (94.4% vs. 98.4%; P < 0.001), and that of eggs laid on the tiers of the aviary was higher in brown compared to white hens (3.12% vs. 0.22%; P < 0.001) (Table 4). Egg production and position of laid eggs did not change over the five weeks of trial (Table 4). Finally, no significant interaction was recorded between enrichment, genotype and/or age on egg production, with the exception of the rate of broken eggs (significan probability of the interaction week of age x enrichment) (Table 4).

Rheological and physical properties of eggs

The enrichment with HI live larvae did not affect most of the rheological and physical properties of eggs (P > 0.05) (Table 5). Nevertheless, a lower proportion of the shell (8.88% vs. 9.03%; P = 0.017) and a trend for lower albumen pH (8.90 vs. 8.97; P = 0.065) were recorded in eggs of hens enriched with HI live larvae compared to those of the control group (Table 5).

Table 5.

Quality traits of laying hens receiving or not 10% live Hermetia illucens (HI) larvae (10% daily feed intake) as feeding enrichment of two genotypes (brown and white) from 36 to 40 weeks of age.

	Enrichment (E)		Genotype (G)		P value			RMSE
	Control	HI	Brown	White	E	G	E × G
Eggs (n)	120	120	120	120
Whole egg weight (g)	71.5	72.3	71.7	72.0	0.310	0.572	0.105	4.30
Pre-cracks (%)	1.67	2.50	1.67	2.50	0.651	0.651	-	-
Blood spots (%)	12.5	10.8	21.7	1.67	0.688	<0.001	-	-
Meat spots (%)	15.0	13.3	27.5	0.83	0.711	<0.001	-	-
Yolk (%)	24.2	24.0	23.2	24.9	0.430	<0.001	0.288	1.48
Albumen (%)	66.8	67.1	67.7	66.3	0.132	<0.001	0.848	1.53
Shell (%)	9.03	8.88	9.13	8.78	0.017	<0.001	0.008	0.48
Shell thickness (mm)	0.463	0.467	0.475	0.455	0.526	<0.001	0.031	0.033
Shape index (%)	1.29	1.29	1.27	1.32	0.822	<0.001	0.447	0.039
Shell strength (N)	15.8	15.4	15.7	15.5	0.379	0.713	0.136	2.18
Albumen pH	8.97	8.90	8.93	8.94	0.065	0.508	0.994	0.013
Yolk pH	6.23	6.27	6.25	6.25	0.257	0.918	0.865	0.283
Yolk color
L*	51.3	51.5	50.6	52.2	0.687	<0.001	0.482	3.70
a*	13.9	13.9	14.1	13.7	0.955	0.146	0.821	2.22
b*	54.5	55.2	55.1	54.6	0.054	0.666	0.829	9.53

RSME, root mean square error.

Control: not enriched with live Hermetia illucens larvae; HI: receiving live Hermetia illucens larvae (10% daily feed intake) as a feeding enrichment.

As for differences according to genotype, the occurrence of blood spots (21.7% vs. 1.67%; P < 0.001) and meat spots (27.5% vs. 0.83%; P < 0.001) was dramatically higher in eggs from brown hens compared to those from white hens (Table 5). Then, the yolk proportion was lower (23.2% vs. 24.9%), the albumen (67.7% vs. 66.3%) and shell (9.13% vs. 8.78%) proportions higher in eggs laid from brown hens compared to those from white hens (P < 0.001), with shells thicker (0.475 mm vs. 0.455 mm; P < 0.001) in the former compared to the latter hens. Finally, the egg shape index was lower (P < 0.001) in eggs from brown hens compared to those from white hens. As for color traits of the yolk, a lower lightness index was measured in eggs from brown hens compared to those from white hens (L*: 50.6 vs. 52.2; P < 0.001) (Table 5).

Chemical composition of yolk and albumen

The enrichment with HI larvae did not affect proximate and mineral composition of yolks, while albumen moisture content was higher (88.0% vs. 87.6%; P = 0.013) and protein content lower in eggs of hens fed HI larvae group compared to those of the control group (10.5% vs. 10.8%; P = 0.012) (Table 6). As for the fatty acid profile of yolks, the proportions of C14:0 (+153%; P < 0.001), other SFA (+40.1%; P = 0.001), other MUFA (+29.3%; P = 0.002) and total SFA (+1.61%; P = 0.009) were higher in yolk lipids of hens enriched with HI larvae compared to those of control hens (Table 7).

Table 6.

Proximate composition of egg yolk and albumen of laying hens receiving or not 10% live Hermetia illucens (HI) larvae (10% daily feed intake) as feeding enrichment of two genotypes (brown and white) from 36 to 40 weeks of age.

	Enrichment (E)		Genotype		P value			RMSE
	Control	HI	Brown	White	E	G	E × G
Eggsa (n)	20	20	20	20
Yolk
Moisture (%)	46.5	49.4	48.9	46.9	0.144	0.290	0.350	5.96
Protein (%)	17.7	16.7	17.0	17.4	0.127	0.426	0.391	1.94
Fat (%)	33.0	31.4	31.6	32.8	0.176	0.325	0.421	3.79
Ash (%)	1.75	1.72	1.64	1.82	0.735	0.065	0.446	0.294
Albumen
Moisture (%)	87.6	88.0	87.7	87.9	0.013	0.223	0.137	0.503
Protein (%)	10.8	10.5	10.8	10.5	0.012	0.124	0.125	0.456
Ash (%)	0.709	0.694	0.709	0.695	0.551	0.564	0.114	0.075

RSME: root mean square error.

^aPools of two yolks or two albumens from the same experimental group.

Control: not enriched with live Hermetia illucens larvae; HI: receiving live Hermetia illucens larvae (10% daily feed intake) as a feeding enrichment.

Table 7.

Fatty acid (FA) composition of egg yolk (% total FA) of laying hens receiving or not 10% live Hermetia illucens (HI) larvae (10% daily feed intake) as feeding enrichment of two genotypes (brown and white) from 36 to 40 weeks of age.

	Enrichment (E)		Genotype (G)		P value			RMSE
	Control	HI	Brown	White	E	G	E × G
Poolsa (n)	20	20	20	20
C14:0	0.338	0.857	0.628	0.566	<0.001	0.471	0.483	0.269
C16:0	25.6	25.9	25.2	26.3	0.106	<0.001	0.392	0.500
C18:0	10.8	10.5	10.1	11.2	0.247	0.002	0.844	0.832
C20:0	0.061	0.056	0.053	0.064	0.714	0.455	0.738	0.044
Other SFA	0.451	0.632	0.593	0.489	0.001	0.119	0.955	0.206
C16:1n9	0.517	0.518	0.617	0.418	0.954	<0.001	0.549	0.070
C16:1n7	2.28	2.38	2.47	2.19	0.291	0.003	0.302	0.274
C18:1n9	30.7	29.9	31.1	29.6	0.131	0.004	0.941	1.55
C18:1n7	1.41	1.42	1.54	1.29	0.710	<0.001	0.137	0.084
Other MUFA	0.389	0.503	0.481	0.411	0.002	0.013	0.331	0.085
C18:2n6. LA	19.6	19.7	19.5	19.7	0.724	0.599	0.774	1.21
C18:3n6. GLA	0.157	0.155	0.142	0.169	0.742	0.002	0.983	0.021
C20:4n6	3.34	3.19	3.13	3.39	0.233	0.039	0.476	0.390
C22:5n6	0.473	0.463	0.482	0.454	0.773	0.445	0.896	0.112
Other PUFA n6	0.903	0.845	0.824	0.924	0.133	0.013	0.182	0.120
C18:3n3. ALA	0.865	0.830	0.873	0.822	0.413	0.227	0.635	0.132
C22:5n3. DPA	0.244	0.279	0.310	0.213	0.139	0.002	0.431	0.074
C22:6n3. DHA	1.88	1.86	1.93	1.81	0.820	0.183	0.596	0.279
Other PUFA n3	0.005	0.004	0.005	0.004	0.756	0.796	0.852	0.012
SFA	37.3	37.9	36.6	38.6	0.009	<0.001	0.730	1.64
MUFA	35.3	34.8	36.2	33.9	0.314	<0.001	0.698	1.68
PUFA	27.4	27.3	27.2	27.5	0.790	0.451	0.841	1.24
PUFA n6	24.4	24.4	24.1	24.7	0.812	0.117	0.923	1.12
PUFA n3	3.00	2.98	3.12	2.85	0.806	0.003	0.592	0.263
PUFA n6/PUFA n3	8.24	8.23	7.78	8.68	0.985	<0.001	0.673	0.660

RSME. root mean square error.

^aPools of two yolks or two albumens from the same experimental group.

Control: not enriched with live Hermetia illucens larvae; HI: receiving live Hermetia illucens larvae (10% daily feed intake) as a feeding enrichment.

As for the effect of the genotype, the rates of C16:0 (+4.37%; P < 0.001), C18:0 (+10.9%; P = 0.002), C18:3n6 (+19.0%; P = 0.002), C20:4n6 (+8.31%; P = 0.039) and other PUFA n6 (+12.1%; P = 0.013) were lower in eggs from brown hens compared to those from white hens, whereas the rates of C16:1n9 (+47.6%; P < 0.001), C16:1n7 (+12.8%; P = 0.003), C18:1n9 (+5.07%; P = 0.04), C18:1n7 (+19.4%; P < 0.001), other MUFA (+17.0%; P = 0.013) and C22:5n3 (+45.5%; P = 0.002) was higher in brown eggs compared to white eggs (Table 7).

Thus, on the whole, the rate of SFA (P < 0.001) and the ratio PUFA n6/n3 (P < 0.001) was lower in brown eggs compared to white eggs, that of MUFA (P < 0.001) and PUFA n3 (+9.47%; P = 0.003) was higher in brown eggs compared to white eggs (Table 7).

As for the mineral composition of the eggs, some differences were recorded according to HI enrichment and genotype (Table 8). In detail, the enrichment with HI larvae decreased the Na content in the yolk (P = 0.008) and albumen (P = 0.007) as well as the S content in albumen (P = 0.003) compared to eggs of not enriched hens. As for the effect of the genotype, eggs from white hens showed higher yolk contents of Mg (P = 0.052) and Na (P = 0.003) compared to brown hens.

Table 8.

Mineral composition of the egg yolk and albumen (mg/kg) of laying hens receiving or not 10% live Hermetia illucens (HI) larvae (10% daily feed intake) as feeding enrichment of two genotypes (brown and white) from 36 to 40 weeks of age.

	Enrichment (E)		Genotype (G)		P value			RMSE
	Control	HI	Brown	White	E	G	E × G
Poolsa (n)	20	20	20	20
Yolk
Ca	1582	1484	1491	1574	0.200	0.274	0.682	237
Fe	59.8	56.9	58.5	58.2	0.399	0.940	0.858	10.6
K	1150	1067	1122	1096	0.131	0.634	0.587	170
Mg	136	127	125	138	0.223	0.052	0.993	21.3
Na	546	497	478	564	0.008	0.003	0.699	84.9
P	6066	5674	5659	6081	0.181	0.151	0.972	909
S	1931	1806	1814	1923	0.182	0.243	0.878	289
Zn	39.1	36.2	36.0	39.3	0.128	0.090	0.957	6.02
Albumen
Ca	91.8	96.7	91.8	96.8	0.422	0.407	0.910	19.0
Fe	0.095	0.093	0.074	0.113	0.933	0.083	0.695	0.069
K	1393	1355	1344	1403	0.355	0.153	0.232	127
Mg	120	114	116	118	0.069	0.453	0.139	9.00
Na	1877	1773	1846	1804	0.007	0256	0.596	114
P	137	134	135	136	0.478	0.913	0.111	13.2
S	1933	1822	1873	1883	0.003	0.778	0.055	111
Zn	0.044	0.048	0.044	0.048	0.318	0.279	0.343	0.014

RSME. root mean square error.

^aPools of two yolks or two albumens from the same experimental group.

Control: not enriched with live Hermetia illucens larvae; HI: receiving live Hermetia illucens larvae (10% daily feed intake) as a feeding enrichment.

Sensory analysis of eggs

At the triangular test, only 21 out of 67 panelists (44.3%) were able to correctly discriminate the egg from hens fed with live HI larvae (data not shown in tables). Then, out of 32 participants who were given eggs from white hens, only 8 of them identified the different one, while out of 35 participants who were given eggs from the brown hens, 13 of them gave the correct answer (data not shown in tables). Differences in correct responses according to HI enrichment and genotype were not statistically significant (P = 0.678), i.e., the number of correct responses did not exceed chance and no significant differences were detected between the samples.

Effect of diet on feces microbiota

The taxonomic classification revealed that Firmicutes was the dominant phylum in the feces microbiota of laying hens, followed by Actinobacteriota and Bacteroidota. Out of the various genera, Brachybacterium was consistently predominant across all samples, while Salinococcus, Yaniella, Lactobacillus, Jeotgalicoccus, Bacteroides, and Brevibacterium were present as minor genera (Fig. 6).

Fig. 6.

Relative abundance of microbial taxa at phylum and genus levels in the feces of laying hens after five weeks of trial: brown-feathered Control hens not receiving the enrichment (BC); white-feathered Control hens not receiving the enrichment (WC); brown-feathered hens enriched with live Hermetia illucens larvae (BI; 10% daily feed intake as feeding enrichment from 36 to 40 weeks of age); white-feathered hens enriched with live Hermetia illucens larvae (WI).

Based on the results of the Shannon index, Simpson index, and PCoA analysis, no significant differences were observed as microbial richness, diversity, and specific taxonomic biomarkers among brown and white-feathered hens, receiving or not the enrichment with HI live larvae (Fig. 7).

Fig. 7.

Beta diversity levels of microbial populations (principal coordinate analysis (PCoA) plot using a Bray-Curtis distance matrix) (a); and alpha diversity as for Shannon index (b) and Simpson index (c) in the feces of laying hens after five weeks of trial: brown-feathered Control hens not receiving the enrichment (BC); white-feathered Control hens not receiving the enrichment (WC); brown-feathered hens enriched with live Hermetia illucens larvae (BI; 10% daily feed intake as feeding enrichment from 36 to 40 weeks of age); white-feathered hens enriched with live Hermetia illucens larvae (WI).

Discussion

The present trial was designed to evaluate the effect of the enrichment with live larvae in two genotypes of laying hens kept under controlled conditions which could be consistent with commercial ones on a small scale. This implied some limitations, such as the number of replications (i.e. 2 pens per experimental group; 4 pens per main experimental factor). Nevertheless, the number of animals per pen (i.e. 225 hens) allowed us to obtain relevant information about hen behaviour and welfare and egg quality.

Animal welfare: behaviour and reactivity of laying hens

Access to live larvae is highly attractive for poultry as they are used as rewards for training these animals in different tests, such as judgement bias tests, preference and motivation tests (Hernandez et al., 2015; Skalná et al., 2023). Whether and how their supplementation can affect the overall welfare of laying hens under farming conditions is less studied yet and has to be evaluated considering both negative and positive welfare indicators. In fact, the provision of live larvae could be expected to promote a positive affective state in laying hens, where positive welfare has been recently defined as “the animal flourishing through the experience of predominantly positive mental states and the development of competence and resilience” (Rault et al., 2025). While animal-based measures for negative welfare in poultry have been widely identified and validated (EFSA, 2023a, 2023b), research on reliable and valid positive welfare indicators is still under development (Paulović et al., 2024).

In our trial, the enrichment with live HI larvae favored species-specific behaviors on the floor area, such as pecking (both insects and feeders), where we also observed a reduction of eggs laid on the floor compared to control pens. In other words, in cage-free systems, the functional use of the different areas has been improved as the higher activity of hens on the floor likely made nesting in the same area somewhat uncomfortable and disturbed, promoting the use of available nests for egg laying, which is positive from a production point of view. Indeed, a positive effect of enrichment on space use has been previously found in pullets because of an early habituation to varied stimuli, helping the use of designated laying areas in adulthood, i.e., increasing the use of large nest boxes and reducing floor eggs (Bari et al., 2020).

As for species-specific behaviors, previous studies also reported that the enrichment with live larvae increased the expression of ground pecking, scratching, and active foraging (Bongiorno et al., 2024) possibly reducing feather pecking, thus contributing to improved plumage condition (Star et al., 2020). Indeed, foraging and exploratory behaviors (e.g., ground pecking, scratching) are highly motivated behaviors, considered potential indicators of a positive affective state in laying hens and broiler chickens (Papageorgiou et al., 2023). Comfort behaviors, like preening, wing stretching, and synchronized resting and dustbathing, have been also associated with relaxed emotional states (Lourenço da Silva et al., 2021; Rasmussen et al., 2024). Indeed, under our conditions, no difference in the number of animals performing dustbathing was recorded between pens enriched or not with HI live larvae. Nevertheless, in laying hens belonging to local breeds (Bionda Piemontese and Bianca di Saluzzo), the occurrence of positive behaviors (preening, allopreening, and walking) significantly increased and severe feather pecking decreased in response to the supplementation with live HI larvae (Bellezza Oddon et al., 2024).

Based on the definition of positive welfare (Rault et al., 2025), animals subjected to positive experiences are likely to have a reaction indicative of higher stress resilience towards new challenging events because of their expected increased competence. In this context, reactivity tests evaluating the response of poultry to challenging events, such as exposure to new/unknow objects/environment and/or humans, can also provide useful information about the animal resilience and affective status (Zulkifli and Azah, 2004; Loconsole and Regolin, 2023), as investigated in our study using the novel object test and the avoidance distance test.

Under our conditions, results from the two tests did not provide a clear picture of the resilience of hens following enrichment with live larvae. In fact, findings in the two tests were inconsistent, without differences in the novel object test and a decreased avoidance distance in white hens receiving the enrichment compared to the other experimental groups (significant interaction Genotype × Enrichment; Fig. 4). Thus, based on our results, we cannot definitively state whether the enrichment with live larvae promoted positive welfare as i) the validity of the indicators that we used for assessing positive welfare has not yet been definitively stated; and ii) we could both hypothesize that 1) white hens were so willing to get live larvae that they approached humans waiting for live larvae during the avoidance distance test; or 2) the administration of live larvae improved their affective state and reduced their fear towards humans. While the absence of major effects on behaviors (with special reference to dustbathing as comfort positive behavior) and the lack of consistency in the response at the two tests would support the first hypothesis, previous studies also provided not consistent results as for the effect of the enrichment with live larvae on the affective status of poultry. In fact, the enrichment with live larvae did not affect behavior in the open field test in Bovan White laying hens (Tahmantani et al., 2021), whereas it reduced fear towards humans with a shorter avoidance distance in medium-growing broiler female chickens, which was not confirmed in male birds, however (Bongiorno et al., 2024). On the other hand, time spent foraging increased both in fast-growing (Ipema et al., 2020) and in medium-growing broiler chickens (Bongiorno et al., 2024) following provision of HI live larvae.

Other studies showed that a complex rearing environment offering different enrichments can reduce fearfulness towards novel stimuli in laying hens (Dumontier et al., 2022) but can also increase feather corticosterone levels in slow-growing broiler chickens (Lourenço da Silva et al., 2023). In other words, some enrichments or ways of offering them can promote stress in animals, even if to a different extent depending on interactions with genotype. As for live larvae, while arousal during delivery may occur as larvae are consumed in a very short time (within 5 minutes; present trial), the effects of live larvae provision remain largely beneficial on negative behaviors, as recorded also under our conditions. In fact, no consistent evidence of increased aggressive or negative social behaviors has been reported following regular provisioning of live larvae (Biasato et al., 2022), whereas feather pecking and stereotypies have been found to decrease in groups enriched with live larvae (Bellezza Oddon et al., 2024). Automatic systems for distributions of live larvae in feeders and/or on the ground (Dörper et al., 2023; Golfidis et al., 2024) or ad libitum provision could reduce any stress associated to a concentrated provision and could differently impact animal behaviors and welfare, which deserves further investigation also in view of differences in temperaments between genotypes.

In fact, significant genotype-related variations have been documented for spatial cognition and, therefore, space and nest use (Farkas et al., 2022; Ciarelli et al., 2023) Moreover, the two genotypes show a different temperament where white hens are usually more nervous and less confident towards humans compared to brown ones (Rentsch et al., 2023; Gulabrai et al., 2025).

Overall, white hens exhibit higher reactivity and fear-related behaviors, showing longer latency to the approach in the novel object tests and longer tonic immobility (underlying heightened fearfulness) compared to brown hens (Rentsch et al., 2023).

Despite we did not observe differences in the occurrence of aggressions between the two genotypes, social dynamics are also genotype dependent: white hens have been found to display more dominant and aggressive behaviors in social groups compared to brown hens (Nie et al., 2019) where visual cues, such as plumage color, can influence social hierarchy and behavioral interactions. In our study, as for negative behaviors, white laying hens showed a greater occurrence of piling behaviors compared to brown hens. In fact, this behavior is a poorly understood group-level behavior, with multifactorial origins. Piling episodes have been related to a variety of environmental and management factors, including inadequate ventilation, high temperatures, sudden changes in light or noise, and high stocking densities (Mazocco et al., 2024; Chowdhury et al., 2025; Gray et al., 2025). While some studies have explored associations between fearfulness, social dynamics, and piling (Mazocco et al., 2024), the interplay between these variables remains unclear, and, to our knowledge, no literature information is available about its relationships with genotypes.

Production and egg quality

Different levels of supplementation with HI live larvae over the diet can have different effects on performance, product quality and welfare of animals. As for performance, in laying hens kept in individual cages (Tahamtani et al., 2021) or in an aviary system (Star et al., 2020), previous studies have tested supplementations between 10% and 20% of daily feed intake which have been proven to be effective in maintaining or enhancing productive performance without adversely affecting body weight or feed consumption. In Isa Brown hens that had access to a free-range area, the free access to HI whole dried larvae over the commercial diet did not impact their live weight (Ruhnke et al., 2018). According to the Authors, the high temperatures that the animals experienced during the summer season of the trial might have reduced their feed intake and, consequently, any effect of the additional nutrient intake through dry whole larvae was observed.

When hen performance is not affected, egg macroscopic quality is not likely to change substantially which is confirmed by the results of above-mentioned studies (Tahamtani et al., 2021; Veldkamp et al., 2024) which found negligible effects of a HI live larvae supplementation (5% to 20% feed intake) on egg rheological and physical traits, proportions of different parts (shell, yolk and albumen), or the Haugh index, consistently with our results.

When HI live larvae were provided ad libitum, Bovans white hens exhibited a significant increase in body weight and a concurrent reduction in the intake of the complete diet, which suggest a potential for a partial dietary substitution (Tahamtani et al., 2021). Nevertheless, even under these conditions, rheological and physical characteristics of eggs did not change, while the yolk lightness index was higher compared to eggs of hens not supplemented or receiving 10-20% HI live larvae (Tahamtani et al., 2021). On the other hand, a higher yolk redness index was measured in eggs obtained from Lohmann Brown Classic laying hens when HI larvae meal fully replaced soybean meal (100% substitution of soybean meal) in the complete diet for 21 weeks, due to a higher γ-tocopherol, lutein, β-carotene, and total carotenoids contents (Secci et al., 2018). In fact, insect supplementation can affect the yolk color to a different extent depending on insect species and product type (characterized by different xanthophyll types and contents), besides supplementation level and period. Under our conditions, the xanthophylls intake through HI live larvae was likely not sufficient to produce any change in the color of yolks with respect to the basal xanthophylls supplementation provided in the commercial diet (capsanthin 0.7 mg/kg, zeaxanthin 0.3 mg/kg, lutein 4.8 mg/kg), despite a suitable supplementation period. On the other hand, Tahamtani et al. (2021) recorded a decrease of yolk color score based on Roche color fan over time in laying hens fed live larvae ad libitum compared to those receiving 10-20% live larvae, which was attributed to a reduction of the carotenoid intake through diet.

As for the chemical composition and nutritional value of eggs, HI live larvae supplementation can be associated with a negative shift in the FA profile of products towards saturated fatty acids due to the prevalence of lauric, palmitic, and myristic acids in HI larvae (Marín et al., 2024), which is confirmed in our study with live larvae rich in SFA (63% of total fatty acids) and C12:0 the most represented (41.3% total FA) (see Table 7). A previous study (Cattaneo et al., 2025) found that a 15% and 30% dietary supplementation/enrichment of HI live larvae to Lohmann Brown hens significantly increased the proportion of SFA (+3.4% and +7.9%, respectively with 15% and 30% HI live larvae) and PUFAs (+15.5% and +17.8%) while decreased the proportion of MUFA (−6.0% and −10.0%) in yolk. Nevertheless, changes were in a small range (6-10%) and slightly impacted on the nutritional value and nutritional indexes for human consumption. These results were confirmed under our conditions, where a 10% supplementation of HI live larvae only produced a slight, even if significant, increase in total SFA (+1.61%), which implies limited effects on the nutritional value of the final products.

As for the sensory traits, under our conditions, no difference was detected by the panelists on eggs from hens receiving or not HI live larvae. Consistently, trained panelists did not discriminate flavor and odor between eggs from hens fed diets with 0%, 5% and 10% of HI live larvae replacing soybean meal (Veldkamp et al., 2024). Differently, in caged hens from an Arabian strain fed diets supplemented with HI meal at 1% and 5%, non-trained panelists found that egg appearance, consistency and flavor improved in eggs laid by hens fed with increasing insect meal rates, without smell modifications (Al-Qazzaz et al., 2016). These Authors attributed the improvement in the egg taste in response to the insect meal dietary inclusion to the high content of glutamic acid of HI larvae. Indeed, different compounds play a role in the flavor and odor of eggs and eggs products, such as volatile compounds and amino acid composition (Goto et al., 2021; Nishimura et al., 2021) and fatty acid composition. This latter one can be affected by the feeding regime, the housing systems and rearing conditions, besides specific interaction with the genotype used (Sirri et al., 2018a; Gautron et al., 2022).

Previous studies showed that in Bovans Brown laying hens, the dietary inclusion of HI larvae meal (5%, 10%, and 15%) increased the calcium serum concentration compared with the control group without effect in eggs traits (Zawisza et al., 2023). On the other hand, the replacement (59%) of soya bean meal with HI larvae meal in a prepupae stage was found to increase the eggshell thickness of eggs because of a higher mineral intake (Kawasaki et al., 2019). Indeed, effects on egg quality are not likely to occur when live larvae are given as an enrichment as the mineral content (with special emphasis on calcium) is higher in the commercial diets compared to the supplemented HI larvae, consistently with what measured in our trial. Then, a few changes in the yolk and albumen contents of other minerals (Na, S,), which, despite being significant, have a low nutritional value (because of their rather low absolute contents). Some changes in the albumen content of Fe, Cr, Zn, and Al have been previously recorded in Lohmann laying hens at 48 weeks of age and fed a diet in which HI larvae meal (included at 21% of the diet) fully replaced soybean meal (Khan et al., 2024).

The genotype of laying hens may influence egg production and egg quality traits, and, thus, economic parameters, interacting with hen age, diet, and housing system (Sirri et al., 2018a; Gautron et al., 2022, 2022).

On the other hand, differences in temperament of laying hens according to genotype may also be associated with differences in egg quality as for the occurrence of specific defects (Simons et al., 2017). In fact, the stress experiences of laying hens, measured in terms of some physiological and behavioral indicators, has been related to the internal egg inclusion of blood and meat spots (Campo and Garcia, 1998). Indeed, under our conditions, a higher occurrence of these defects was measured in the eggs of brown hens compared to those of white hens, which is aligned with previous studies (Wu et al., 2025) but not consistent with the differences we observed in the reactivity of brown and white hens at the avoidance distance test and novel object test. Overall, causes for blood or meet spots may be several and related to the diet (insufficient dietary intake of vitamins A and K, excessive dietary inclusion rate of lucerne meal, mold contamination) as well as challenging management conditions (e.g., continuous light program, panic movements) (Simons et al., 2017).

On the other hand, we did not record any difference between the two commercial genotypes tested in the chemical composition of eggs, as for the content of the main nutrients, which is known to be rather stable and not affected by main ontogenetic, management and feeding conditions within similar genetic lines (Onbaşılar et al., 2018). More differences can be expected when comparing selected crossbred genotypes with pure local breeds with consequences on the functional properties of eggs (because of the protein components) and their nutritional properties (because of the fatty acid profile) for human health (Cartoni Mancinelli et al., 2022b; Gautron et al., 2022).

Under our conditions, the fatty acid profile, of eggs of white hens showed a higher proportion of SFA and a lower proportion of MUFA and PUFA n-6 compared to eggs of brown hens. Indeed, across different commercial genotypes and local breeds, several studies have reported genotype-specific profiles in the relative proportions of SFA, MUFA, and PUFA (Sirri et al., 2018b; Hejdysz et al., 2024), as well as in the proportions of the most important PUFA series (Sirri et al., 2018b), i.e., n-3 and n-6 series. These differences have been related to breed-specific metabolic pathways and different liver synthesis of fatty acids starting from dietary carbohydrates (Cartoni Mancinelli et al., 2022a; Hejdysz et al., 2024). On the other hand, previous results suggest that eggs of local breeds have a healthier lipid profile in the yolk compared to commercial genotypes as for the lower n-6/n-3 ratio (Sirri et al., 2018b), which is expected to be relatively more beneficial to human health by reducing the risk of chronic inflammatory diseases (Dal Bosco et al., 2024).

Fecal microbiota

In our trial, fecal microbiota changes according to larvae provision were also measured as a possible welfare indicator related to the gut health of laying hens (Bellezza Oddon et al., 2024; Huang et al., 2024). Nevertheless, we did not find any effect on gut microbial composition, likely because of both the relatively low supplementation rate (10% of feed intake) and the type of supplementation (live larvae). In fact, other studies found that the dietary inclusion of insect meal both in broiler chickens (0.2% and 0.3% Tenebrio molitor and Zophobas morio larvae meal) (Józefiak et al., 2020) and in laying hens (7.5% of Tenebrio molitor larvae meal) (Biasato et al., 2018) have been associated with an improved balance between beneficial and pathogenic bacteria. In Xuefeng black-bone hens fed 1% to 5% HI larvae meal, a richer microbial population and an improved health status have been found (He et al., 2021). Overall, the dominance of Firmicutes, followed by Actinobacteriota and Bacteroidota, that we observed in the feces of laying hens, consistently aligns with previous studies indicating their central role in poultry feces microbiomes as fermentable bacteria group (Cressman et al., 2010; Bindari et al., 2021; Horyanto et al., 2024). Then, the prevalence of Brachybacterium across all samples also suggests a ubiquitous presence of this taxon in poultry gut (Valeris-Chacin et al., 2021), possibly due to its capacity to degrade organic matter.

Different genotypes can also be expected to display different gut microbial profiles, which in turn can affect their resilience towards health challenges and, in turn, their welfare. Indeed, we did not observe any difference between genotype on the taxonomy and diversity feces microbial populations, whereas previous studies have shown that brown and white feathered strains Lohmann hens can exhibit distinct cecal microbiome, with the former showing lower diversity, decreased abundance of Lactobacillus, and an increase in genera such as unclassified Enterobacteriaceae and Escherichia-Shigella compared with the latter (Lu et al., 2025). Clear differences in microbial diversity and taxonomic abundance were also found between Hy-Line Gray hens and Lohmann Pink hens, whereas both genotypes showed the same dominant phyla, i.e., Bacteroidetes, Firmicutes, and Fusobacteria, followed by Proteobacteria, Euryarchaeota, Melainabacteria, and unidentified bacteria (all > 1% relative abundance) (Huang et al., 2019).

Conclusions

The enrichment with HI live larvae, at the supplementation rates tested in the present study, can be successfully used in laying hens kept in cage-free aviary systems to increase species-specific behaviors without negatively or substantially affecting egg production and/or egg physical or nutritional quality. At the same time, the use of space is improved as eggs in the nests increase at the expense of floor eggs, which is finally translated into an improvement of production. As for hen welfare according to live larvae enrichment, the improved acceptance towards humans at the avoidance distance test can be positively interpreted as a reduction of fear and an improvement of the hen affective states. On the other hand, further investigations are needed as for the effect on gut microbiota, which absence in our study could be attributed to the samples used (feces), the larvae supplementation rate, and/or the common environment in which hens were kept. Specific genotype-responses to enrichment should be further considered to tailor management in view of animal welfare and farm production optimization under the different production systems.

Data availability statement

Sequencing data have been deposited in the European Nucleotide Archive (ENA - European Bioinformatic Institute, Hinxton, UK) under the project accession number: PRJEB9618.

CRediT authorship contribution statement

M. Pravato: Writing – review & editing, Writing – original draft, Visualization, Methodology, Investigation, Formal analysis, Data curation. C. Ciarelli: Writing – review & editing, Investigation, Conceptualization. A. Miolo: Writing – review & editing, Investigation, Conceptualization. F. Bordignon: Writing – review & editing, Investigation, Formal analysis. J. Vas: Writing – review & editing, Visualization. G. Zardinoni: Writing – review & editing, Methodology, Investigation. L. Giagnoni: Writing – review & editing, Investigation. A. Squartini: Writing – review & editing, Methodology. P. Stevanato: Writing – review & editing, Resources. G. Xiccato: Writing – review & editing, Supervision, Resources, Funding acquisition, Conceptualization. A. Trocino: Writing – review & editing, Writing – original draft, Supervision, Resources, Project administration, Investigation, Funding acquisition, Conceptualization.

Disclosures

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.