Effects of coarse corn or oat hulls on growth performance, intestinal health, and microbiota modulation in underperforming broilers

Original Research Article Effects of coarse corn or oat hulls on growth performance, intestinal health, and microbiota modulation in underperforming broilers Muhammad Zeeshan Akram a , b , Ester Ar� evalo Sureda a , Matthias Corion a , Luke Comer a , Haoran Zhao a , Martine Schroyen b , Nadia Everaert a , * a Nutrition and Animal-Microbiota Ecosystems Laboratory, Department of Biosystems, KU Leuven, Heverlee 3000, Belgium b Precision Livestock and Nutrition Unit, Gembloux Agro-Bio Tech, University of Li� ege, Gembloux, Belgium article info Article history: Received 6 January 2025 Received in revised form 9 April 2025 Accepted 16 April 2025 Available online 6 August 2025 Keywords: Coarse corn Oat hulls Underperforming broiler Gut health Microbiota abstract Intra-flock body weight (BW) variability in broilers increases production costs, as underperforming chicks often show suboptimal gut development and performance. Increasing grain particle size and dietary fiber content has been shown to improve digestive efficiency and intestinal health. This study investigated whether dietary inclusion of coarse corn (CC) and oat hulls (OH) could improve gut health and reduce the performance gap between lowand high-BW (LBW and HBW) broilers. On d 7, 1400 Ross 308 male broilers were categorized as LBW or HBW, with 504 LBW chicks assigned to 4 isocaloric and isonitrogenous diets with 10% fine corn (LBWC), 7% CC with 3% fine corn (LBW +CC), 3% OH with 10% fine corn (LBW +OH), or 7% CC and 3% OH (LBW +CO). High BW chicks received a 10% fine corn diet (HBWC). Each group had 6 replicates with 21 chicks per pen. The HBWC group showed the highest BW at each timepoint (P < 0.05). By d 38, LBW +OH chicks had significantly reduced the weight difference with HBWC chicks and significantly outperformed LBWC chicks (P < 0.001), whereas other groups showed intermediate values. Coarse corn and OH, individually or combined, reduced the relative plasma FITC-dextrann concentration d 14 (P =0.014) and increased gizzard weights on d 21 and 38 (P < 0.05) as compared with LBWC group. The LBW +OH group showed increased pancreas relative weight on d 21 (P =0.005, vs. HBWC) and villus height (P =0.042, vs. LBWC) on d 38. Additionally, LBW +OH group reduced isobutyrate and isovalerate levels in cecum (P < 0.05, vs. HBWC and LBWC) on d 21, and upregulated ileal genes related to gut barrier function (CLDN1, vs. HBWC and LBWC; CLDN4, vs. HBWC; CLDN5, vs. LBWC), amino acid and glucose transporters (SLC15A1 and SLC1A4, vs. HBWC and LBWC), and immune function (NOS2, vs. HBWC and LBWC; TLR4, vs. HBWC) on d 14 (P < 0.05), and sodiumphosphate transporter SLC34A2 (P =0.049, vs. HBWC) on d 38. LBW +CC birds upregulated SLC15A1 (vs. HBWC and LBWC) on d 38 (P < 0.001). Lactobacillus was enriched in the cecum of HBWC birds, while Escherichia-Shigella was abundant in LBWC birds on d 14, with CC and OH promoting beneficial bacterial shifts in LBW groups. Overall, incorporating structural components into diets, particularly 3% OH, enhanced gastrointestinal development, intestinal integrity, and growth performance in LBW broilers. These improvements reduced disparities in BW between LBW and HBW birds, thereby contributing to more uniform flock performance at slaughter age. © 2025 The Authors. Publishing services by Elsevier B.V. on behalf of KeAi Communications Co. Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 1. Introduction The broiler industry confronts persistent challenges in managing weight heterogeneity, particularly for low body weight (LBW) broilers that consistently underperform compared to their normal-weight counterparts. These birds represent a significant economic and production challenge, characterized by suboptimal performance, health issues and welfare concerns (Hughes et al., 2017). Low body weight broilers are uniquely vulnerable, *Corresponding author. E-mail address: [email protected] (N. Everaert). Peer review under the responsibility of Chinese Association of Animal Science and Veterinary Medicine Production and Hosting by Elsevier on behalf of KeAi Contents lists available at ScienceDirect Animal Nutrition journal homepage: http://www.keaipublishing.com/en/journals/aninu/ https://doi.org/10.1016/j.aninu.2025.04.012 2405-6545/© 2025 The Authors. Publishing services by Elsevier B.V. on behalf of KeAi Communications Co. Ltd. This is an open access article under the CC BY license (http:// creativecommons.org/licenses/by/4.0/). Animal Nutrition 23 (2025) 153–166 exhibiting profound physiological limitations that extend beyond mere growth restrictions. These limitations include severe gastrointestinal tract (GIT) developmental impairments, such as increased intestinal permeability and the downregulation of gut barrier function proteins and nutrient transporters (Akram et al., 2024c; Zhang et al., 2022). Furthermore, LBW broilers often harbor dysbiotic microbiota, leading to elevated proinflammatory cytokine levels and increased intestinal inflammation (Akram et al., 2024a,b,c; Zhang et al., 2022), which further compromises their growth and health. Optimal broiler growth and production fundamentally depend on efficient nutrient digestion and absorption, which are intrinsically linked to a well-functioning digestive system (El Sabry and Yalcin, 2023). In recent years, nutritional strategies aiming to modulate gut structure and function have gained prominence, particularly the incorporation of dietary insoluble fiber sources and coarse feed particles (Jim� enez-Moreno et al., 2013; Xu et al., 2015). Adding insoluble fiber sources such as oat hulls (OH) has been found to improve nutrient retention, digestion, and growth in broilers (Naeem et al., 2023). Likewise, the addition of coarse corn (CC) to pelleted diets enhances protein digestibility, energy utilization, live performance, and litter quality (Xu et al., 2015). These benefits are attributed to the physical properties of these feed ingredients, which stimulate gizzard development (Sarbast K Kheravii et al., 2017), and increase pancreatic enzyme secretion, such as amylase and chymotrypsin, driven by gizzard activity. Furthermore, a well-developed gizzard promotes the release of cholecystokinin (CCK), which stimulates reverse peristalsis, prolonging digesta transit time and enabling more thorough digestion (Denbow, 2015; Svihus et al., 2004; Svihus, 2011). Consequently, slower digesta transit increases nutrient digestion and absorption by maximizing the contact time with absorptive cells (Washburn, 1991). Besides the chemical composition and particle size, the physical structure of feed including pellet size and hardness may influence digestive development. Harder pellets can further stimulate gizzard activity by resisting breakdown, thereby enhancing the mechanical stimulation of the digestive tract (Abdollahi and Ravindran, 2013). While the breakdown of fiber in poultry is minimal in terms of energy provision, it may still influence the nutritional value of feed through interactions with other nutrients. Unlike soluble fiber, which can hinder nutrient digestion and absorption due to increased digesta viscosity (Smits et al., 1997), insoluble fiber supports chicken growth by improving the nutrient digestibility in the upper GIT and stimulating microbial fermentation in the lower tract (Naeem et al., 2023). However, the effects of insoluble fiber can be context-dependent. At inappropriate inclusion levels or depending on the fiber source and bird physiology, insoluble fiber may impair nutrient utilization or lead to undesirable outcomes such as wet litter or sticky droppings, often considered antinutritional effects in broiler production systems (Jha and Mishra, 2021). The physical attributes of OH and CC may also exert microbiota-modulating effects. The ligninrich matrix of fiber materials just as in OH can act as a fermentable substrate for beneficial microbes (Kheravii et al., 2017), producing short-chain fatty acids (SCFAs) with anti-inflammatory and trophic effects on the gut epithelium. Similarly, the structural complexity of CC may support the proliferation of specific bacterial taxa that favor gut health and metabolic efficiency (Yan et al., 2022). While individual studies have explored the effects of CC and OH on nutrient digestibility and growth performance, their impact on gut microbiota and intestinal health is relatively underexplored. The positive effects of coarse grain particles and insoluble fiber on broiler performance are well documented (Sarbast K Kheravii et al., 2017; Kheravii et al., 2018a). However, their potential benefits in LBW broilers, which are characterized by impaired growth performance and physiological development, remain elusive. Given the unique physiological deficits in LBW birds, such interventions may offer greater relative benefits by compensating for early developmental delays. Therefore, this study aimed to evaluate the individual and combined effects of dietary CC and OH on growth performance, GIT morphology, intestinal health markers, and microbiota characteristics in LBW broilers. We hypothesized that the inclusion of these structural dietary components would stimulate gizzard activity, improve gut function and microbial balance, thereby enhancing growth and narrowing the performance disparity between LBW and high body weight (HBW) broilers. 2. Materials and methods 2.1. Animal ethics statement The study was performed at TRANSfarm, KU Leuven, Bierbeek, Belgium, following approval from the KU Leuven Ethical Committee for Animal Experimentation under project number 112/ 2023. 2.2. Experimental diets Isocaloric and isonitrogenous wheat-based broiler diets were formulated to meet nutritional requirements across different growth phases (Table 1). All broilers received a crumbled-form pre-starter diet in the first week. Thereafter, 4 pelleted experimental diets were formulated in a commercial feed mill (Vanden Avenne Commodities, Ooigem, Belgium) using conditioning with expander: a commercial broiler diet with 10% finely ground corn (control), a diet formulated with 7% CC and 3% finely ground corn (CC), a diet containing 10% finely ground corn and 3% OH (OH), and a combination diet with 7% CC and 3% OH (CO). The grower diets were fed until d 16, followed by the finisher diets provided until the end of the trial. Fine corn, wheat, and soybean meal were ground using a hammer mill fitted with a 4-mm sieve, while coarse corn was processed using a roller mill with sequential gap settings of 1.8, 1.6, and 1.5 mm. Oat hulls, initially pelleted, were reground using a roller mill with fixed gaps of 3.6 mm. All experimental diets were pelleted using a die with 3.2-mm holes and a roll-die distance of 0.2 mm. Pelleting involved an expander time of approximately 5 s at 20 bar (2.0 MPa) and 80 ◦ C, followed by a conditioning phase at 65 ◦ C for 10 s with 2.1% steam addition and 12% initial feed moisture. The particle size distribution of the OH was assessed through dry sieving in duplicate. Results indicated that 62% of particles were greater than 4.00 mm, 10% between 3.15 and 4.00 mm, 13% between 2.00 and 3.15 mm, 11% between 1.00 and 2.00 mm, and 4% < 1.00 mm. The particle size distribution (%), geometric mean diameter (GMD) and geometric standard deviation (GSD) of the experimental feeds were determined through wet sieving in duplicate (Table 2). A 20-g feed sample was soaked in 400 mL of distilled water for 1 h at room temperature. The sample was then sieved using a vibratory sieve shaker (Retch NV, Aartselaar, Belgium) equipped with sieves with mesh sizes of 2000, 1000, 500, 200, 90, 50, and 38 μm. The feed and water suspension were deposited onto the top sieve, and sieving was performed for 10 min with a water flow rate of 2.0 to 2.3 L/min, followed by 1 min without water flow to drain excess moisture. The fractions retained on each sieve were collected separately in Falcon tubes, freeze-dried, and stored in a desiccator until weighing. The M.Z. Akram, E.A. Sureda, M. Corion et al. Animal Nutrition 23 (2025) 153–166 154 Table 1 Composition and nutrient level of the wheat-based diet (%, DM basis). Item Pre-starter (d 1 to 7) Grower (d 8 to16) Finisher (d 17 to 38) Control CC OH CO Control CC OH CO Ingredients Wheat 43.48 52.93 52.92 49.35 49.35 55.89 55.89 53.47 53.47 Soyabean meal 28.21 26.16 26.17 28.48 28.46 21.84 21.84 24.69 24.67 Coarse corn 0.00 0.00 7.00 0.00 7.00 0.00 7.00 0.00 7.00 Oat hull 0.00 0.00 0.00 3.00 3.00 0.00 0.00 3.00 3.00 Fine corn 15.00 10.00 3.00 10.00 3.00 10.00 3.00 10.00 3.00 Soya oil 4.77 4.83 4.83 5.49 5.49 4.83 4.83 5.08 5.08 Sunflower meal 3.57 2.40 2.40 0.00 0.00 3.90 3.90 0.00 0.00 Monocalcium phosphate 1.29 0.89 0.89 0.92 0.92 0.93 0.93 0.97 0.97 Limestone 1.26 0.93 0.93 0.92 0.92 0.84 0.84 0.82 0.82 Salt 0.20 0.22 0.22 0.22 0.22 0.21 0.21 0.22 0.22 Na-bicarbonate 0.20 0.17 0.17 0.16 0.16 0.18 0.18 0.17 0.17 Choline 75% 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 Lysine 0.66 0.48 0.48 0.47 0.49 0.50 0.50 0.49 0.50 Methionine 0.36 0.29 0.29 0.30 0.30 0.26 0.26 0.28 0.28 L-Valine 0.12 0.06 0.06 0.06 0.06 0.05 0.05 0.06 0.06 L-Isoleucine 0.07 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.02 L-Arginine 0.06 0.00 0.00 0.00 0.00 0.00 0.00 0.02 0.03 L-Threonine 0.21 0.15 0.15 0.14 0.14 0.14 0.14 0.30 0.30 Vitamin E 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 Vitamin premix 1 0.39 0.31 0.31 0.31 0.31 0.30 0.30 0.30 0.30 Decoquinate 2 0.05 0.05 0.05 0.05 0.05 0.00 0.00 0.00 0.00 Phytase 3 0.008 0.008 0.008 0.008 0.008 0.008 0.008 0.008 0.008 Xylanase 4 0.00 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 Nutrient levels 5 Metabolizable energy, kcal/kg 2975 2925 2925 2925 2925 2950 2950 2950 2950 Crude protein 21.30 20.20 20.18 20.17 20.21 19.00 19.09 19.02 19.04 Crude fat 5.92 5.23 5.46 5.82 5.93 5.66 5.68 6.08 5.84 Crude fiber 3.18 3.07 3.06 3.58 3.57 3.22 3.21 3.56 3.55 NDF 8.37 9.32 9.11 17.7 15.46 9.01 8.45 12.83 11.38 ADF 4.61 4.58 4.83 5.01 5.41 4.37 4.21 4.95 4.49 ADL 0.63 0.62 0.62 0.70 0.68 0.69 0.67 0.71 0.73 Digestible lysine 1.22 1.08 1.05 1.10 1.08 1.00 1.01 1.04 1.01 Calcium 0.7 0.69 0.69 0.72 0.73 0.69 0.73 0.69 0.68 Phosphorus 0.70 0.57 0.51 0.59 0.56 0.59 0.60 0.57 0.57 Sodium 0.14 0.13 0.13 0.14 0.14 0.15 0.14 0.14 0.14 Chloride 0.20 0.21 0.19 0.22 0.20 0.19 0.21 0.23 0.20 Potassium 0.94 0.90 0.88 0.94 0.93 0.85 0.85 0.87 0.88 Magnesium 0.21 0.20 0.21 0.18 0.19 0.18 0.17 0.19 0.18 Control =a commercial broiler diet with 10% finely ground corn; CC =a diet formulated with 7% coarse corn and 3% finely ground corn; OH =a diet containing 10% finely ground corn and 3% oat hulls; CO =a diet containing 7% coarse corn and 3% oat hulls. 1 Premix provided per kilogram of diets: vitamin A 10,000 IU, vitamin D 3 2750 IU, 25-hydroxycho2ecalciferol 0.056 mg, vitamin E 90 mg, copper 15 mg, iron 15 mg, manganese 85 mg, zinc 50 mg, iodine 2 mg, and selenium 0.4 mg. 2 Provided per kilogram of diets: 30.3 mg of decoquinate. 3 Provided per kilogram of diets: 500 FTU. 4 Provided per kilogram of diets: 10 IU. 5 Metabolizable energy was calculated, while all other nutrient levels were analyzed. Table 2 Wet sieving particle size distribution (%) and geometric mean diameter (GMD) of experimental diets used in pre-starter, grower and finisher phases for lowand high-weight broilers. Item Pre-starter (d 1 to 7) Grower (d 8 to 16) Finisher (d 17 to 38) Control CC OH CO Control CC OH CO Sieve size 2.000 mm 1.3 1.0 10.1 5.4 7.4 4.7 13.7 5.4 15.5 1.000 mm 6.1 7.6 16.8 8.4 14.1 9.4 18.9 11.4 25.1 0.500 mm 10.2 9.1 12.1 13.1 12.7 13.0 12.2 11.7 10.0 0.200 mm 12.7 11.0 8.4 9.8 10.6 9.7 7.6 8.7 6.3 0.090 mm 10.2 9.4 5.1 8.4 5.3 8.4 4.6 7.5 3.8 0.050 mm 10.2 7.2 6.1 6.4 6.4 6.4 5.5 5.7 4.5 0.038 mm 6.3 9.4 3.0 8.4 3.2 8.4 2.7 7.5 2.3 <0.038 mm 43.1 45.3 38.4 40.2 40.3 40.1 34.7 42.0 32.6 >1.000 mm 7.4 8.6 26.9 14.0 21.5 14.1 32.6 16.8 40.6 GMD ± GSD 360 ±54.1 372 ±48.7 582 ±57.6 416 ±51.5 532 ±82.2 413 ±45.9 621 ±108.2 458 ±53.3 698 ±101.2 Control =a commercial broiler diet with 10% finely ground corn; CC =a diet formulated with 7% CC and 3% finely ground corn; OH =a diet containing 10% finely ground corn and 3% OH; CO =a diet containing 7% CC and 3% OH; GSD =geometric standard deviation. M.Z. Akram, E.A. Sureda, M. Corion et al. Animal Nutrition 23 (2025) 153–166 155 average particle size (d av ) was calculated according to the equation: d av =∑d i ×m i ; where d i represents the mesh size of sieve i, and m i represents the mass percentage of the fraction retained on sieve i. Except stated otherwise, all chemical analyses were done using AOAC (2016) methods. Crude protein was analyzed by the Kjeldahl method (method 990.03), and crude fat was via Soxhlet extraction (method 920.39). Crude fiber was analyzed using the fiber bag technique (method 978.10). Neutral detergent fiber (NDF) and acid detergent fiber (ADF) were analyzed following the procedures described by Van Soest et al. (1991), while acid detergent lignin (ADL) was measured following method 973.18. Digestible lysine levels were calculated based on the analyzed amino acid profile, analyzed using high-performance liquid chromatography following method 982.30. Calcium, phosphorus, sodium, potassium, and magnesium were quantified using inductively coupled plasma optical emission spectrometry following method 985.01. Chloride content was analyzed using potentiometric titration following method 943.01. 2.3. Animals, husbandry, and data collection A total of 1400 one-day-old male Ross 308 broiler chicks (initial BW: 44.5 ±3.21 g) were obtained from a commercial hatchery (Belgabroed NV, Aartselaar, Belgium). All chicks were fed a prestarter diet until d 7, after which individual BW was recorded, and the birds were categorized into low-, medium-, and highweight groups on the basis of BW distribution. Day 7 BW is a strong predictor of final BW, and selection at this age enables biologically relevant classification of LBW and HBW phenotypes before compensatory growth mechanisms begin to manifest, which typically occur during later stages of development (Akram et al., 2024c,b). The chicks on the lower end of the BW range were designated as LBW (n =504), and further divided into 4 treatment groups: LBWC, LBW +CC, LBW +OH, and LBW +CO, each comprising 126 chicks receiving either the control diet or one of the experimental diets (CC, OH, or CO). Additionally, the birds on the higher end of the BW (HBWC, n =126) received the control diet with fine corn, whereas the remaining medium BW chicks (n =770) were excluded from the experiment. The experiment consisted of 5 groups with 6 replicate pens per group (1.3 m 2 per pen), with each pen containing 21 birds. Pen floors were covered with 3 cm of wood shavings. Feed and water were provided ad libitum throughout the study. The light schedule started with 1 h of darkness on d 1, increasing by 1 h per day to 6 h of darkness, which was maintained thereafter. The initial temperature was set at 33 ◦ C and decreased by 0.5 ◦ C daily until it reached 21.5 ◦ C on d 21, after which it remained constant. 2.4. Sampling and measurements BW was individually recorded, and feed intake was recorded per pen after each dietary phase. The average daily gain (ADG), mortality-corrected average daily feed intake (ADFI) and feed conversion ratio (FCR) were calculated for the grower and finisher phases. The birds weighing nearest to the pen’s average weight (n =12) were sacrificed through electrical stunning followed by decapitation on d 14, 21, and 38. Dissection was performed, and the weights of the gizzard, liver, pancreas, small intestine, and cecum were determined. The small intestine and cecum weights were measured without emptying the digesta, and their length was also recorded. The relative organ weights were expressed as g/ 100 g BW, and the relative lengths of the small intestine and cecum were calculated as cm/100 g BW. Digesta samples from both caeca were collected, placed in 2-mL vials, snap-frozen in liquid nitrogen, and stored at −80 ◦ C for microbiota and volatile fatty acid (VFA) analysis. Ileum sections from the midpoint were taken for histomorphological examination on d 14, 21 and 38, while ileal tissue samples were snap-frozen and stored at −80 ◦ C for highthroughput qPCR gene expression analysis on d 14 and 38. Two chickens per pen in each group were randomly selected on d 14, 21 and 38 for intestinal permeability tests via fluorescein isothiocyanate (FITC)-dextran (4000 kDa; Sigma–Aldrich, St. Louis, MO, USA). 2.5. Ileal histomorphology Ileum samples were fixed in 4% formaldehyde for 48 h and afterwards stored in 70% ethanol. Histology sections were embedded in paraffin, sectioned, and stained with alcian blueperiodic acid–Schiff according to the standard procedure of the GIGA immunohistochemistry platform (ULi� ege, Belgium). The microscopy images were analyzed via NDP.view2 software (Hamamatsu Photonics K.K., Hamamatsu, Japan). Villus height (VH) and crypt depth (CD) were measured for 20-well-oriented villus-crypt units per bird, and the VH/CD ratio was calculated. 2.6. Intestinal permeability FITC-dextran solution (2.2 mg/mL per bird) was administered orally, and blood samples (1 mL) were collected from the jugular vein 2.5 h post gavage. Blood samples were centrifuged at 4 ◦ C at 3000×g. A standard series was created, and plasma samples diluted in phosphate-buffered saline (1:5) were analyzed in duplicate via a 96-well microplate reader (CLARIOstar Plus, BMG LABTECH GmbH, Offenburg, Germany) with an excitation wavelength of 485 nm and an emission wavelength of 530 nm. Plasma FITC-dextran concentrations (ng/mL) were calculated using a standard curve. The relative concentration of FITC-dextran was calculated as ng/mL per 100 g BW. 2.7. Cecal volatile fatty acid analysis Short-chain fatty acids (acetate, propionate, butyrate, valerate, and caproate) and branched-chain fatty acids (BCFAs: isobutyrate, isovalerate, and isocaproate) were measured according to a modified method from previous study (Van Craeyveld et al., 2008). Briefly, approximately 250 mg of cecal content was weighed into 2-mL Eppendorf tubes, placed on ice, and mixed with 50 μL of MHA-2 internal standard solution and 80 μL of 6 mol/L HCl. The samples were vortexed and were left on ice for 20 min and then mixed with 25% NaCl and tertiarybutyl methyl ether. After centrifugation at 4 ◦ C (10,000 ×g for 5 min), 600 μL of the supernatant was transferred to a 1.5-mL Eppendorf tube containing anhydrous sodium sulfate, vortexed, and centrifuged again. A 200-μL aliquot was pipetted into screw-neck vials with conical glass inserts and stored at −20 ◦ C until analysis. Volatile fatty acids were quantified via an HP 6890 Series GC System equipped with an automatic liquid sampler, flame ionization detector, and DB-FFAP capillary column (30 m length, 0.32 mm internal diameter, 0.25 μm film thickness; Agilent Technologies, Santa Clara, CA, USA). The carrier gas was nitrogen at a flow rate of 25 mL/min, with the column at 130 ◦ C and the injector and detector at 195 ◦ C. SCFA and BCFA concentrations were calculated in mmol/g wet digesta on the basis of calibration curves. M.Z. Akram, E.A. Sureda, M. Corion et al. Animal Nutrition 23 (2025) 153–166 156 2.8. Gene expression through high-throughput qPCR 2.8.1. Selection of genes, primer design and validation A total of 92 genes (13 housekeeping genes and 79 target genes) involved in intestinal barrier function, nutrient transport, immune response, metabolism, and oxidative homeostasis were selected based on published literature (Akram et al., 2024c, 2025). Details on the genes and their primary functions are provided in Table S1. Primers were designed using the NCBI Primer-Blast tool to span exon–exon junctions to minimize genomic DNA amplification. Specificity was confirmed by melting curve analysis following qPCR amplification, which revealed single peaks for all primers, indicating that there was no non-specific amplification or primerdimer formation. Verification using agarose gel electrophoresis revealed single and distinct bands at the expected molecular weights for each amplicon. The primer efficiency was optimized between 90% and 110%, with R 2 values exceeding 0.99, using 3-fold serial dilutions of pooled cDNA derived from all samples on a QuantStudio 6 Real-Time PCR System (Thermo Fisher Scientific, Waltham, MA, USA). 2.8.2. RNA extraction Total RNA was extracted from ileal tissue samples using the ReliaPrep RNA Miniprep System (Promega Corporation, Madison, WI, USA) according to the manufacturer’s protocol. The RNA concentration and purity were determined by spectrophotometry (Nanodrop 2000, Thermo Fisher Scientific, Waltham, MA, USA), whereas the integrity of the RNA was verified on 1% agarose gel electrophoresis. 2.8.3. Reverse transcription, preamplification, and high-throughput qPCR The BioMark HD system (Standard BioTools, South San Francisco, CA, USA) was used for high-throughput qPCR, following a protocol described in our previous study (Akram et al., 2024c, 2025). cDNA synthesis was performed using RT MasterMix (Standard BioTools, South San Francisco, CA, USA). Preamplification was conducted in a 96-well qPCR plate with a primer mixture and PreAmp Mastermix (Standard BioTools, South San Francisco, CA, USA) under the following thermal conditions: 95 ◦ C for 2 min, then 14 cycles of 95 ◦ C for 15 s and 60 ◦ C for 4 min. Exonuclease I treatment was then applied to remove unincorporated primers, and pre-amplified cDNA samples were diluted 8fold. Prior to high-throughput qPCR, we prepared a sample mixture by combining 2.25 μL of exonuclease-treated, pre-amplified cDNA with 2.5 μL of 2 ×SSoFast EvaGreen Supermix (BioRad, Hercules, CA, USA) and 0.25 μL of 20 ×DNA-binding dye (Standard BioTools). The assay mixture contained 0.5 μL of each primer (100 μmol/L), 2.5 μL of 2 ×Assay Loading Reagent (Standard BioTools), and 2.25 μL of low-EDTA DNA suspension buffer (TEKnova, Hollister, CA, USA). Both sample and assay mixtures were loaded into 96.96 Integrated Fluid Circuits (IFCs), and qPCR was conducted with initial denaturation at 95 ◦ C for 60 s, followed by 30 cycles of 96 ◦ C for 5 s and 60 ◦ C for 20 s. The standard curve, generated from dilutions of the pooled pre-amplified cDNA, was used to calculate relative mRNA levels. Four reference genes (TBP, B2M, NDUFA, and β-actin) were identified as the most stable under the experimental conditions using the NormFinder algorithm (Andersen et al., 2004). Relative gene expression was calculated using the Pfaffl method (Pfaffl, 2001) through normalization of target genes to the geometric mean of the reference genes' expression levels. Genes showing poor amplification efficiency, high quantification cycle (Cq) variability between technical replicates, or mean Cq values above 30 were excluded from downstream statistical analyses to ensure reliability of the expression data. 2.9. DNA extraction, 16S rRNA gene amplicon sequencing and bioinformatics Cecal digesta DNA was extracted using the QIAamp PowerFecal Pro DNA Kit (Qiagen Benelux B.V., Venlo, Netherlands) following the manufacturer’s protocol. DNA concentration and quality were assessed following the same protocol as described in the RNA extraction section. For sequencing library preparation, the V3–V4 region of the 16S rRNA gene was amplified using primers 341F (5 ′ - CCTAYGGGRBGCASCAG-3 ′ ) and 806R (5 ′ -GGACTACNNGGGTATCTAAT-3 ′ ), each with sample-specific barcodes. Libraries were sequenced on an Illumina NovaSeq 6000 platform, which produced 250 bp paired-end reads. Ultrapure water was included as a negative control to monitor sequencing quality. The raw sequences were subjected to quality filtering, trimming, and demultiplexing in QIIME2 (v2024.2) with the default settings. Low-quality reads were removed, and amplicon sequence variants (ASVs) were generated using DADA2. Taxonomic assignment of ASVs was conducted with the Naïve Bayes classifier against the SILVA database (release 138) at a 99% similarity threshold. For statistical analysis, the QIIME2 artifacts were imported into R (v4.2.3, R Foundation, Vienna, Austria). Microbial diversity metrics (Shannon and Simpson indices) were calculated as alpha diversity measures in R using the phyloseq package (v1.40.0) after rarefaction to the minimum sample depth, and groups were compared via the Kruskal–Wallis test. Beta diversity was evaluated with Bray–Curtis dissimilarity and visualized through principal coordinate analysis (PCoA). Group differences were assessed using non-parametric permutational multivariate analysis of variance (PERMANOVA) with 9999 permutations (vegan package, v2.6.4). Differential microbial abundance at the genus level was determined using linear discriminant analysis effect size (LEfSe) using the microbiome package (v1.18.0), with a linear discriminant analysis (LDA) score threshold of ≥2.0 and significance set at P < 0.05. False discovery rate (FDR) adjustment was applied using the Benjamini-Hochberg method (Benjamini and Hochberg, 1995), and the results were visualized as log 10 (LDA score) values. 2.10. Statistical analysis Data normality was assessed using the Shapiro–Wilk test in R before conducting statistical analyses. Outliers were identified and removed based on values exceeding quartile 3 +1.5 ×interquartile range or falling below quartile 1 −1.5 ×interquartile range. Growth performance, digestive organ characteristics, ileal histomorphology, intestinal permeability, and ileum gene expression were analyzed using a one-way ANOVA model as follows: Y ij =μ+T i +ϵ ij ; where Y ij is the dependent variable; μ is the overall mean; T i is the fixed treatment effect; ϵ ij is the residual error term. Pairwise comparisons were conducted using Tukey’s HSD test. P < 0.05 was set significant difference and 0.05 < P ≤0.10 as a tendency. For gene expression data, P-values were adjusted for FDR using the Benjamini–Hochberg method. All the results are reported as the means with a pooled standard deviation (SD), which combines the variability observed in all samples. Principal component analysis (PCA) was performed to visualize sample clustering on the basis of gene expression data using the factoextra package (v1.0.7) in R. The multivariate effects of the treatments on sample clustering were tested using PERMANOVA with the adonis2 function (v2.6.4) to test for multivariate effects of dietary treatments on sample clustering in PCA. Heatmaps were plotted to M.Z. Akram, E.A. Sureda, M. Corion et al. Animal Nutrition 23 (2025) 153–166 157 show the sample variability and gene expression levels using the pheatmap package (v1.0.12) in R. Two-way hierarchical clustering of the heatmap data was performed using Pearson’s correlation distance and Ward’s clustering methods, with gene expression levels scaled per gene. 3. Results 3.1. Growth performance On d 7 and 16, the BW of the HBWC group was higher than that of all LBW groups (P < 0.05; Table 3). Among LBW groups on d 16, the LBW +OH group had higher BW than the LBWC group (P < 0.001), LBW +CC chicks and LBW +CO chicks had intermediate weights. By d 38, the HBWC group maintained the highest BW, exceeding the LBWC group (P < 0.001). LBW +OH birds showed a significantly reduced difference in BW compared to HBWC birds and had significantly higher BW than LBWC birds (P < 0.001), while LBW +CO and LBW +CC birds showed intermediate values. From d 8 to 16, the HBWC group had higher ADG than all other groups (P < 0.001). Among the LBW groups, the LBW +OH group exhibited greater ADG than the LBWC group (P < 0.001). Over the full study period (d 8–38), HBWC birds maintained the highest ADG, significantly surpassing LBWC, LBW +CC, and LBW +CO birds (P =0.001). The LBW +OH group attained the highest ADG among all LBW groups, significantly exceeding the LBWC group (P =0.001). From d 8 to 16, the HBWC group had higher ADFI than all LBW groups (P =0.023). Over the entire study period, HBWC birds had the highest ADFI, significantly surpassing LBWC and LBW +CO birds (P =0.005). LBW +CC and LBW +OH birds showed intermediate ADFI values that did not differ significantly from HBWC or LBWC birds. 3.2. Relative lengths and weights of digestive organs On d 21, the LBW +OH group had the highest relative pancreas weight, exceeding that of the HBWC group (P =0.005, Table 4). By d 38, the LBW +CO group showed a lower relative liver weight compared to the HBWC group (P =0.027), whereas the other groups had liver weights comparable to HBWC birds. The gizzard relative weight was higher in LBW +CC, LBW +OH, and LBW +CO birds than in both HBWC and LBWC birds on d 21 and 38 (P < 0.05, respectively). 3.3. Ileal histomorphology On d 21, HBWC birds had greater VH compared to LBWC and LBW +CC birds (P =0.005, Table 5). VH in LBW +OH and LBW +CO birds was intermediate, significantly exceeding LBW +CC (P =0.005) but comparable to HBWC birds. By d 38, LBW +OH birds showed the highest VH, significantly surpassing LBWC birds (P =0.042). On d 14, CD was lowest in LBW +CC compared to LBWC, LBW +OH, and LBW +CO birds (P =0.017). 3.4. Intestinal permeability On d 14, 21, and 38, absolute plasma FITC-dextran concentrations did not differ significantly among groups (P > 0.05, Table 6). On d 14, relative plasma FITC-dextran levels were lower in the LBW +CC, LBW +OH, and LBW +CO groups compared to the LBWC group (P =0.014) and were comparable to the HBWC group. On d 38, LBWC birds had a significantly higher relative plasma FITC-dextran concentration than HBWC birds (P =0.022), while LBW groups fed CC, OH or their combination had intermediate values that were not significantly different from those of the HBWC group. 3.5. Cecal VFA composition Isobutyrate tended to be lower in the LBW +CO group on d 14 (P =0.091, Table 7). On d 21, valerate concentration was highest in the LBW +CC group, significantly exceeding that of the LBW +OH group (P =0.002). Isobutyrate, isovalerate and total BCFAs on the same day were significantly lower in the LBW +OH group compared to the LBWC group (P < 0.05, respectively). 3.6. Ileal gene expression 3.6.1. Principal component analysis and heatmap clustering The PCA on d 14 and 38 showed no distinct clustering of experimental groups (Fig. S1). This was supported by PERMANOVA, which revealed no significant associations between Table 3 Effects of coarse corn and oat hulls on the growth performance of low-weight broilers 1 . Item Day Groups SD P-value HBWC LBWC LBW +CC LBW +OH LBW +CO BW, g 7 203.3 a 164.5 b 166.6 b 165.9 b 164.5 b 16.45 <0.001 16 675.8 a 583.9 c 598.5 bc 611.7 b 602.3 bc 61.12 <0.001 38 3087 a 2796 b 2928 ab 2983 a 2941 ab 105.4 <0.001 ADG, g/d 8 to 16 52.5 a 46.6 c 47.6 bc 49.5 b 48.6 b 2.45 <0.001 17 to 38 109.6 100.6 105.9 107.9 106.5 5.36 0.086 8 to 38 80.1 a 72.4 c 76.1 b 77.3 ab 76.3 b 4.31 0.001 ADFI, g/d 8 to 16 71.7 a 56.2 b 57.6 b 58.1 b 56.8 b 9.98 0.023 17 to 38 146.1 134.7 142.7 140.0 137.4 7.40 0.062 8 to 38 124.5 a 111.9 b 118.0 ab 116.2 ab 114.0 b 6.60 0.005 FCR, g/g 8 to 16 1.22 1.20 1.20 1.19 1.17 0.161 0.735 17 to 38 1.33 1.33 1.33 1.31 1.29 0.069 0.850 8 to 38 1.34 1.31 1.31 1.29 1.27 0.070 0.608 Values with different superscripts in a row differ at P < 0.05. HBWC =high BW chickens fed a commercial broiler diet with 10% finely ground corn; LBWC =low body weight chickens fed a commercial broiler diet with 10% finely ground corn; LBW +CC =low body weight chickens fed a commercial broiler diet with 7% coarse corn and 3% finely ground corn; LBW +OH =low body weight chickens fed a commercial broiler diet with 10% ground corn and 3% oat hulls; LBW +CO =low body weight chickens fed a commercial broiler diet with 7% coarse corn and 3% oat hulls. BW =body weight; ADG =average daily gain; ADFI =average daily feed intake; FCR =feed conversion ratio; SD =standard deviation. 1 The BW was measured individually using the animal as the experimental unit. The ADG ADFI and FCR were calculated from 6 replicates of each group using the pen as an experimental unit. M.Z. Akram, E.A. Sureda, M. Corion et al. Animal Nutrition 23 (2025) 153–166 158 Table 4 Effects of coarse corn and oat hulls on the relative weights (g/100 g BW) and lengths (cm/100 g BW) of the digestive organs of low-weight broilers. Item Day Groups SD P-value HBWC LBWC LBW +CC LBW +OH LBW +CO Pancreas, g 14 0.38 0.35 0.39 0.39 0.35 0.097 0.733 21 0.29 b 0.34 ab 0.33 ab 0.37 a 0.33 ab 0.045 0.005 38 0.16 0.17 0.18 0.16 0.17 0.040 0.745 Liver, g 14 2.89 3.06 3.05 2.91 2.86 0.393 0.703 21 2.72 2.77 2.87 2.78 2.79 0.265 0.758 38 2.51 a 2.35 ab 2.34 ab 2.38 ab 2.21 b 0.023 0.027 Gizzard, g 14 2.05 1.89 2.01 2.09 1.97 0.223 0.220 21 1.60 b 1.55 b 1.78 a 1.83 a 1.82 a 0.259 0.016 38 0.90 b 0.93 b 1.01 a 1.05 a 0.99 a 0.191 0.044 Small intestine, g 14 7.97 7.90 7.80 7.45 7.15 0.907 0.137 21 11.49 7.62 7.22 6.98 6.98 6.475 0.374 38 5.18 4.92 5.36 4.69 5.21 0.748 0.215 Cecum, g 14 0.72 0.83 0.78 0.70 0.78 0.207 0.560 21 0.84 0.89 0.79 0.92 0.91 0.228 0.627 38 0.69 0.66 0.75 0.71 0.76 0.191 0.688 Small intestine, cm 14 23.69 25.93 25.71 24.70 23.69 3.173 0.089 21 15.33 15.81 16.36 15.99 15.94 1.387 0.339 38 6.82 7.17 7.35 7.01 7.55 0.871 0.273 Cecum, cm 14 3.85 4.37 5.01 4.04 4.02 0.506 0.160 21 2.64 2.89 2.85 2.91 2.86 0.342 0.291 38 1.27 1.46 1.38 1.42 1.50 0.230 0.143 HBWC =high BW chickens fed a commercial broiler diet with 10% finely ground corn; LBWC =low body weight chickens fed a commercial broiler diet with 10% finely ground corn; LBW +CC =low body weight chickens fed a commercial broiler diet with 7% coarse corn and 3% finely ground corn; LBW +OH =low body weight chickens fed a commercial broiler diet with 10% ground corn and 3% oat hulls; LBW +CO =low body weight chickens fed a commercial broiler diet with 7% coarse corn and 3% oat hulls. SD =standard deviation. Values with different superscripts in a row indicate a significant difference (n =12, P < 0.05). Table 5 Effects of coarse corn and oat hulls on the villus height, crypt depth and villus height to crypt depth of low-weight broilers. Item Day Groups SD P-value HBWC LBWC LBW +CC LBW +OH LBW +CO Villus height, μm 14 548.8 541.3 525.3 542.5 544.9 53.90 0.863 21 704.9 a 633.9 bc 605.8 c 663.1 ab 653.3 abc 68.35 0.005 38 955.5 ab 881.8 b 929.1 ab 998.9 a 951.8 ab 95.77 0.042 Crypt depth, μm 14 148.0 ab 161.2 a 142.0 b 163.6 a 163.9 a 20.73 0.017 21 168.9 164.4 157.2 161.4 158.6 19.42 0.608 38 139.6 131.4 140.1 132.9 139.6 21.26 0.769 Villus height/crypt depth ratio 14 3.72 3.40 3.76 3.36 3.36 0.517 0.123 21 4.26 3.88 3.87 4.14 4.18 0.557 0.284 38 6.90 6.82 6.79 7.64 6.90 1.008 0.201 HBWC =high BW chickens fed a commercial broiler diet with 10% finely ground corn; LBWC =low body weight chickens fed a commercial broiler diet with 10% finely ground corn; LBW +CC =low body weight chickens fed a commercial broiler diet with 7% coarse corn and 3% finely ground corn; LBW +OH =low body weight chickens fed a commercial broiler diet with 10% ground corn and 3% oat hulls; LBW +CO =low body weight chickens fed a commercial broiler diet with 7% coarse corn and 3% oat hulls. SD =standard deviation. Values with different superscripts in a row indicate a significant difference (n =12, P < 0.05). Table 6 Effects of coarse corn and oat hulls on plasma absolute and relative fluorescein isothiocyanate (FITC) dextran levels in low-weight broilers. Item Day Groups SD P-value HBWC LBWC LBW +CC LBW +OH LBW +CO Absolute plasma FITC-dextran, ng/mL 14 45.7 55.8 46.1 48.2 48.3 10.49 0.119 21 52.2 53.0 51.1 51.4 51.9 4.95 0.931 38 59.7 64.8 64.0 62.1 63.1 5.45 0.429 Relative plasma FITC-dextran, ng/mL per 100 g BW 14 7.27 b 10.05 a 8.17 b 8.29 b 8.03 b 2.027 0.014 21 4.23 4.98 4.72 4.65 4.75 0.593 0.063 38 1.88 b 2.31 a 2.12 ab 2.03 ab 2.08 ab 0.283 0.022 HBWC =high BW chickens fed commercial broiler feed with fine corn; LBWC =low body weight chickens fed a commercial broiler diet with 10% finely ground corn; LBW +CC =low body weight chickens fed a commercial broiler diet with 7% coarse corn and 3% finely ground corn; LBW +OH =low body weight chickens fed a commercial broiler diet with 10% ground corn and 3% oat hulls; LBW +CO =low body weight chickens fed a commercial broiler diet with 7% coarse corn and 3% oat hulls. SD =standard deviation. Values with different superscripts in a row indicate a significant difference (n =12, P < 0.05). M.Z. Akram, E.A. Sureda, M. Corion et al. Animal Nutrition 23 (2025) 153–166 159 principal component variability and gene expression differences on d 38, though a tendency for differentiation was observed on d 14 (P =0.073). Heatmaps visualized gene expression variability across samples from all groups (Figs. S2 and S3). Consistent with PCA, 2-way hierarchical clustering on d 14 and 38 demonstrated that neither samples nor genes grouped consistently by experimental group or functional category. 3.6.2. Differential gene expression To identify differentially expressed genes, one-way ANOVA with Tukey’s HSD post hoc test was performed, considering genes with an FDR-adjusted P < 0.05 as significant (Table 8). On d 14, the LBW +OH group showed the highest expression of the tight junction genes CLDN1, CLDN4, and CLDN5; CLDN1 was even significantly higher than that of the LBWC and HBWC groups (P =0.004), while CLDN4 expression was significantly higher compared to the HBWC group (P =0.003) and CLDN5 was significantly higher compared to the LBWC group (P =0.001). The gut hormone CCK gene was upregulated in LBWC birds compared to LBW +CC and LBW +OH birds (P =0.021), with intermediate levels in HBWC and LBW +CO birds. Expression of the nutrient receptor T1R1 was significantly higher in LBW +OH birds than in all other groups (P =0.011). Among the immune function genes, TLR4 and NOS2 (pattern recognition receptor and antiinflammatory marker, respectively) showed the highest expression in LBW +OH birds (P < 0.05, respectively), although TLR4 was not significantly different from LBWC birds, and NOS2 was not significantly different from LBW +CO birds. The inflammatory marker IL-6 was significantly upregulated in LBW +CC birds relative to HBWC, LBWC, and LBW +OH birds (P =0.017). Genes associated with nutrient transport, including SLC15A1 (peptide transporter) and SLC1A4 (amino acid transporter), were upregulated in LBW +OH birds compared with LBWC and HBWC birds (P < 0.001 and P =0.008, respectively). Additionally, the expression of genes such as SLC2A1 (glucose transporter) and VDR (vitamin D receptor) tended to increase in LBW +OH birds (P > 0.05), whereas the expression of CALB1 (calcium-binding protein) tended to increase in HBWC birds (P =0.081). The expression of HMOX2, a gene associated with intestinal oxidation, was increased in the LBW +OH group (P =0.045). On d 38, CLDN1 expression was significantly upregulated in HBWC birds compared to LBW +CC birds (P =0.027). CLDN2 expression tended to be higher in LBW +CC birds (P =0.079). The expression of the proinflammatory cytokine IL-18, an immunerelated gene, was highest in HBWC birds, significantly exceeding LBW +CC, LBW +OH and LBW +CO groups (P =0.003). Among nutrient transporters, SLC15A1 (peptide transporter) was significantly upregulated in LBW +CC birds (P < 0.001), whereas SLC2A2 (glucose transporter) was highest in HBWC birds (P =0.008). SLC34A2 expression (sodium-phosphate cotransporter) was significantly upregulated in LBW +OH compared to HBWC (P =0.049), while SLC30A1 (zinc transporter) showed a tendency toward higher expression in the LBW +CO group (P =0.084). 3.7. Cecal microbiota Cecal amplicon sequencing generated 15,130,747 reads, with 86,461 ±8181 (mean ±SD) reads per sample. After quality filtering, 13,617,672 reads remained, with an average of 77,815 ±7363 reads per sample. 3.7.1. Core microbiota composition Compositional analysis revealed considerable inter-individual variability and a significant shift in the gut microbiota from d 14 to d 38 post-hatching. On d 14, the Firmicutes phylum dominated (95%–97%), with minor contributions from Bacteroidota (1%–2%), Proteobacteria (1%–3%), and Cyanobacteria (0.2%–1.8%) (Table S2). By d 21, Firmicutes remained predominant (93%–96%), but there were slight increases in Actinobacteriota (0.2%–3.6%). On d 38, the dominance of Firmicutes persisted (92%–95%), while Table 7 Effects of coarse corn and oat hulls on cecal volatile fatty acids (mmol/g wet digesta) in low-weight broilers. Item Day Groups SD P-value HBWC LBWC LBW +CC LBW +OH LBW +CO Acetate 14 250.0 226.8 225.7 265.6 210.4 81.34 0.368 21 246.8 259.1 245.3 255.8 243.1 65.38 0.978 38 244.0 320.0 252.3 277.9 278.3 106.87 0.477 Propionate 14 24.8 20.5 21.4 22.6 20.8 3.75 0.125 21 23.5 25.3 23.6 22.2 21.2 6.23 0.468 38 27.6 31.8 29.8 31.3 27.6 8.98 0.704 Butyrate 14 59.8 48.1 43.1 61.2 44.1 20.33 0.111 21 59.5 56.6 45.5 57.5 53.2 22.59 0.644 38 64.9 81.0 69.2 71.4 75.1 39.03 0.695 Valerate 14 3.3 3.2 3.3 3.6 3.2 0.55 0.564 21 3.41 a 3.34 ab 3.54 a 3.04 b 3.26 ab 0.423 0.002 38 4.24 4.26 4.69 4.39 4.02 1.103 0.535 Total SCFAs 14 320.9 290.5 291.1 352.9 271.9 107.34 0.314 21 333.2 344.3 317.9 338.5 320.6 85.18 0.912 38 340.7 437.1 356.0 385.1 385.0 146.49 0.436 Isobutyrate 14 2.7 3.6 2.9 2.6 2.5 0.94 0.091 21 2.57 ab 2.93 a 2.78 a 2.31 b 2.51 ab 0.351 0.005 38 3.71 3.26 3.83 3.68 3.91 0.986 0.564 Isovalerate 14 2.3 2.8 2.7 2.7 2.4 0.91 0.392 21 2.31 ab 2.63 a 2.41 ab 2.03 b 2.24 ab 0.325 0.003 38 3.41 2.91 3.57 3.29 3.42 0.913 0.479 Total BCFAs 14 4.9 6.1 5.1 5.1 4.9 1.71 0.491 21 4.89 ab 5.56 a 4.94 ab 4.34 b 4.76 ab 0.751 0.029 38 7.12 6.17 7.40 6.97 7.33 1.867 0.496 HBWC =high BW chickens fed commercial broiler feed with fine corn; LBWC =low body weight chickens fed a commercial broiler diet with 10% finely ground corn; LBW +CC =low body weight chickens fed a commercial broiler diet with 7% coarse corn and 3% finely ground corn; LBW +OH =low body weight chickens fed a commercial broiler diet with 10% ground corn and 3% oat hulls; LBW +CO =low body weight chickens fed a commercial broiler diet with 7% coarse corn and 3% oat hulls. SD =standard deviation; BCFAs =branched-chain fatty acids. Values with different superscripts in a row indicate a significant difference (n =12, P < 0.05). M.Z. Akram, E.A. Sureda, M. Corion et al. Animal Nutrition 23 (2025) 153–166 160 Actinobacteriota (1.8%–2.8%) and Cyanobacteria (0.6%–1.2%) remained stable. Additionally, Bacteroidota exhibited minor fluctuations (1%–2%) over time, and low-abundance phyla such as Desulfobacterota (<0.3%) remained consistently present. At the genus level, core genera including Faecalibacterium, Lactobacillus, unclassified Lachnospiraceae, and Ruminococcus torques group accounted for approximately 56% of the total relative abundance in the chickens, as shown in Fig. 1. With age, Lactobacillus and unclassified Lachnospiraceae numerically increased, whereas Faecalibacterium and Ruminococcus torques group decreased. Despite the consistency of these core genera across individuals, many low-abundance genera collectively made up more than 10% of the community, representing a highly variable component of the gut ecosystem. 3.7.2. Alpha and beta diversity On d 14, the Shannon index of alpha diversity was significantly higher in the LBW +CC group compared to LBWC and HBWC groups (P < 0.05, Fig. 2), with the LBW +OH and LBW +CO groups showing intermediate values. No significant differences in alpha diversity metrics were observed among groups on d 21 and 38. Beta diversity analysis using Bray–Curtis dissimilarity did not reveal a clear separation in the overall microbiota composition between groups at any timepoint (Fig. S4). This lack of separation was confirmed by PERMANOVA, which found no significant associations between principal component variability and microbiota composition differences (P > 0.05). 3.7.3. Differential abundance analysis of bacterial genera LEfSe analysis identified differentially abundant bacterial genera across groups on d 14, 21, and 38 (LDA cut-off ≥2.0, FDRadjusted P < 0.05). On d 14, a total of 12 genera were identified across groups (Fig. 3A). The HBWC group was enriched with Lactobacillus, whereas LBWC birds had increased Escherichia-Shigella. The LBW +CC group was enriched with unclassified Oscillospiraceae, Ruminococcaceae UCG-005, Gastranaerophilales, the Christensenellaceae R-7 group, Anaerofilum, unclassified Ruminococcaceae, and Candidatus_Soleaferrea. The LBW +OH group showed a higher abundance of Faecalibacterium and Blautia, whereas the LBW +CO group had higher abundance of unclassified Lachnospiraceae. On d 21, a total of 9 genera were identified (Fig. 3B). HBWC birds were enriched with Anaerostipes, and LBWC birds with UCG-008 (family Butyricicoccaceae) and Colidextribacter. LBW +CC birds showed higher abundance of Ruminococcaceae UCG-005, Roseburia, Romboutsia, and Akkermansia, while LBW +OH birds were enriched with unclassified Lachnospiraceae. LBW +CO birds showed increased Bifidobacterium abundance. On d 38, a total of 13 genera were identified (Fig. 3C). HBWC birds showed a higher abundance of Lactobacillus and Caproiciproducens, while LBWC birds were enriched with Lachnospiraceae_FE2018_group. LBW +CC birds showed enrichment of Christensenellaceae R-7 group and unclassified Ruminococcaceae, Mordavella, and Eggerthellaceae. LBW +OH birds were enriched with Fournierella and Lachnospiraceae_FCS020_group, whereas LBW +CO birds showed increased abundance of Romboutsia, Enterococcus, [Clostridium]_sprioforme and Sellimonas. 4. Discussion The relationship between first-week chick weight and subsequent broiler performance represents a critical determinant of production efficiency, as demonstrated by current findings and supported by previous research (Akram et al., 2024b). Despite uniform rearing conditions, HBWC chicks consistently Table 8 Effects of coarse corn and oat hulls on the relative expression of genes involved in various intestinal functions in the ileum on d 14 and 38 1 . Genes Function Groups SD FDR-adjusted P-value HBWC LBWC LBW +CC LBW +OH LBW +CO Day 14 CLDN1 Barrier function 1.34 b 1.37 b 1.10 b 2.26 a 1.42 b 0.827 0.004 CLDN4 Barrier function 0.54 b 0.98 ab 0.51 b 1.62 a 1.57 a 1.084 0.003 CLDN5 Barrier function 1.58 ab 0.86 bc 0.45 c 1.89 a 0.74 bc 1.019 0.001 CDX Barrier function 0.98 1.05 0.99 1.33 1.39 0.440 0.083 CCK Gut hormone 1.72 ab 2.33 a 1.13 b 0.97 b 1.73 ab 1.052 0.021 T1R1 Nutrient receptor 0.89 b 1.09 b 0.92 b 1.70 a 0.95 b 0.618 0.011 NOS2 Immune function 0.87 b 0.83 b 1.03 b 1.50 a 1.11 ab 0.588 0.044 TLR4 Immune function 0.56 b 0.96 ab 0.52 b 1.14 a 0.67 b 0.758 0.021 IL-6 Immune function 0.24 b 0.22 b 0.90 a 0.32 b 0.56 ab 0.535 0.017 SLC15A1 Nutrient transport 0.64 c 0.62 c 1.63 ab 2.07 a 1.23 bc 0.881 <0.001 SLC1A4 Nutrient transport 1.00 b 1.02 b 0.88 b 1.94 a 1.26 ab 0.752 0.008 SLC2A1 Nutrient transport 0.98 0.85 0.98 1.70 1.45 0.816 0.064 VDR Nutrient transport 1.39 1.20 1.01 1.51 0.59 0.894 0.085 CALB1 Nutrient transport 1.60 1.09 1.20 1.14 1.09 0.500 0.081 HMOX2 Oxidation 0.71 b 0.78 b 0.94 ab 1.17 a 0.94 ab 0.397 0.045 Day 38 CLDN1 Barrier function 0.94 a 0.51 ab 0.33 b 0.69 ab 0.79 ab 0.613 0.027 CLDN2 Barrier function 1.06 1.23 2.02 1.57 1.39 0.859 0.079 IL-18 Immune function 1.40 a 1.09 ab 0.39 c 0.50 bc 0.58 bc 0.695 0.003 TLR4 Immune function 0.42 0.96 0.27 0.46 0.36 0.641 0.094 SLC15A1 Nutrient transport 0.78 b 1.05 b 2.97 a 1.34 b 1.06 b 0.903 <0.001 SLC2A2 Nutrient transport 1.63 a 0.91 b 0.76 b 0.42 b 0.51 b 0.515 0.008 SLC34A2 Nutrient transport 0.45 b 0.89 ab 1.03 ab 1.31 a 0.95 ab 0.814 0.049 SLC30A1 Nutrient transport 1.59 1.52 2.17 1.94 2.22 0.764 0.084 HBWC =high BW chickens fed a commercial broiler diet with 10% finely ground corn; LBWC =low body weight chickens fed a commercial broiler diet with 10% finely ground corn; LBW +CC =low body weight chickens fed a commercial broiler diet with 7% coarse corn and 3% finely ground corn; LBW +OH =low body weight chickens fed a commercial broiler diet with 10% ground corn and 3% oat hulls; LBW +CO =low body weight chickens fed a commercial broiler diet with 7% coarse corn and 3% oat hulls. SD =standard deviation; FDR =false discovery rate. Values with different superscripts in a row indicate a significant difference (n =12, FDR adjusted P < 0.05). 1 Only significantly different or tended to be different genes are shown. M.Z. Akram, E.A. Sureda, M. Corion et al. Animal Nutrition 23 (2025) 153–166 161