Low dietary calcium impairs bone quality and induces bone site-specific remodeling in aged laying ducks with minimal alterations in cecal microbiota and metabolite profiles

Abstract

Calcium (Ca) is critical for poultry health and production, with deficiency impairing skeletal function and productivity. However, the effects of dietary Ca deficiency on the gut–microbiota–metabolite–bone axis in aged laying ducks remain poorly understood. In this study, 144 healthy 60-week-old Longyan Shanma laying ducks were randomly assigned into two groups: a normal-Ca (3.6%, NC) and low-Ca (1.8%, LC) group, and fed for 5 weeks. The LC diet decreased egg production, egg mass and eggshell quality, while elevating feed conversion ratio (P < 0.05). It also reduced bone strength, mineral density and contents in various bone tissues. Serum levels of 1,25-dihydroxyvitamin D₃and procollagen type I carboxy-terminal propeptide (PICP) were markedly increased. The expression of bone metabolism-related genes, including receptor activator of nuclear factor-κB (RANK) and receptor activator of nuclear factor-κB ligand (RANKL), were upregulated in both tibia and femur. Moreover, dietary low Ca damaged intestinal morphology by reducing intestinal villus height and villus height-to-crypt depth (V/C) ratio in the duodenum, jejunum and ileum (P < 0.05). Furthermore, the LC diet reduced cecal microbial alpha diversity, accompanied by altered microbial composition and functional reprogramming involved in carbohydrate metabolism and ion transport. Low calcium also reshaped cecal metabolite profiles, with obvious changes in ursolic acid, galactosylsphingosine and picolinic acid levels. In summary, dietary low calcium disrupts bone homeostasis and intestinal barrier function in aged laying ducks. Importantly, tibia and femur exhibit distinct transcriptional responses to calcium deficiency, revealing evident bone site-specific regulatory patterns, which provides novel insights into the gut-bone axis mechanism underlying calcium deficiency-induced bone metabolic disorders.

Introduction

Bone remodeling plays a pivotal role in poultry health and production by sustaining skeletal growth and development, providing a dynamic calcium reserve for eggshell formation, and alleviating leg problems to improve animal welfare (Fu et al., 2024; Sharma et al., 2023). Recently, the intricate crosstalk between gut microbiota and bone health has increasingly been recognized as a key determinant of skeletal homeostasis (D'Amelio and Sassi, 2018; Wang et al., 2022). Probiotics exert their effects primarily by reshaping the gut microbiota, and this modulation can subsequently impact bone metabolism, for instance by promoting bone turnover and enhancing osteoprotegerin (OPG) expression within bone tissue (Amin et al., 2020). Modulating the gut and its microbiome influences bone density, structure, and strength, which is mainly attributed to three mechanisms: reducing inflammation in the gut, bloodstream, and bone tissue; enhancing calcium absorption and local signaling in the gut-bone axis via metabolites such as short-chain fatty acids (SCFAs); and regulating bone density through bacterial secretory factors and intestinal hormones such as incretins and serotonin (McCabe et al., 2015). Microbiota-derived metabolites are critical for bone homeostasis: SCFAs promote bone formation by regulating insulin-like growth factor-1 (IGF-1), while trimethylamine N-oxide enhances osteogenic differentiation and bone quality by upregulating the expression of osteoblast-related genes (runt-related transcription factor 2 (RUNX2), bone morphogenetic protein-2 (BMP-2) (Han et al., 2024). Collectively, gut microbiota and their derived metabolites can regulate bone remodeling, homeostasis, and quality in poultry.

Calcium (Ca) exerts diverse beneficial effects on the biological, metabolic, and physiological functions of poultry. It plays a vital role in sustaining poultry health and preventing nutritional disorders (Omori et al., 2022; Dongare et al., 2024). Calcium deficiency is associated with impaired growth and reduced productive performance (Li et al., 2020; Hu et al., 2021; Zhang et al., 2020; Ghasemi et al., 2019; Chen et al., 2015a), while decreased bone mineral density directly compromises skeletal strength in laying hens (Bello et al., 2020), and impairs eggshell quality (Zhao et al., 2020), particularly in aged laying poultry. Furthermore, dietary calcium modulates gut microbiota composition in Wistar rats (Fuhren et al., 2021). Supplementing diets with extra calcium phosphate reduced intestinal mucosal damage in rats orally infected with Salmonella enteritidis by regulating intestinal microflora (Bovee-Oudenhoven et al., 1999). In broiler chickens, dietary supplementation with 1.05% active dicalcium phosphate inhibited pathogenic bacteria and increased beneficial gut bacteria in cecal contents (Xing R, 2020).In weaned piglets, diets supplemented with different calcium sources altered the relative abundance of Actinobacteriota at the phylum level, with key differential intestinal microbes primarily involved in carbohydrate and amino acid metabolism, membrane transport, and cofactor and vitamin metabolism (Yang et al., 2022). These findings suggest a relationship between calcium and gut microbiota. However, whether dietary calcium levels can affect bone quality and metabolism by regulating gut microbiota or their metabolites in aged laying ducks remains unknown.

Thus, the current study aimed to investigate the effects of dietary low calcium levels on the productive performance, egg quality, bone quality, metabolic markers, gut morphology, and cecal microflora and metabolites of aged laying ducks. The findings will provide a scientific basis for regulating gut microflora to improve bone and eggshell quality in aged laying ducks.

Materials and methods

Experimental animals, diets, and design

The experimental protocol and use of ducks were approved by the Animal Care and Use Committee of the Animal Science Institute of Guangdong Academy of Agriculture Sciences (Approval No.: 2024002). A total of 144 healthy 60-week-old Longyan laying ducks were randomly assigned to two dietary treatments, with 6 replicates per treatment and 12 ducks per replicate. The treatments consisted of a basal diet with a normal calcium (Ca) level (3.6%, NC) or a low Ca level (1.8%, LC). The composition and nutrient levels of the basal diet are presented in Table 1. Dietary Ca content was determined using a Kjeltec 8400 Analyzer (FOSS Analytical AB, Hoganas, Sweden) following the Chinese National Standard method (GB/T 18868-2002). During the experiment, ducks were individually housed in cages (42 cm × 30 cm × 50 cm) equipped with a feeder and nipple drinker (Guangzhou Huanan Poultry Equipment, Guangzhou, PR China). All ducks had ad libitum access to drinking water and were fed twice daily at 07:00 and 15:00. The feeding trial lasted for 5 weeks. This duration was determined according to previous findings, which confirmed that low dietary calcium could induce detectable changes in bone mineral density and eggshell quality in laying hens within 4–8 weeks (26 to 34 weeks of age; Zhao et al., 2020). In addition, medullary bone calcium mobilization has been proven to be rapidly triggered within four days under low-calcium conditions (0.13% Ca; De Bernard et al., 1980). Moreover, the average egg production of laying ducks fed LC diet was lower than 40% (ranges from 30.6% to 43%) at the 5th week of the trial in our study. We therefore set the trial feeding period at five weeks.

Table 1.

Composition and nutrient levels of the basal diets (air-dry basis).

	Dietary Ca level
Ingredients, %	3.6% (NC)	1.8% (LC)
Corn (CP 7.8%)	49.42	46.31
Wheat bran (CP 15.7%)	18.15	29.41
Soybean meal (CP 43%)	21.18	17.76
Calcium hydrogen phosphate	1.18	1.09
Limestone	8.60	3.91
Salt	0.30	0.30
DL-methionine	0.14	0.14
L-lysine sulfate	0.04	0.08
Premix1	1.00	1.00
Total	100	100
Nutrient levels2
AME, MJ/kg	10.46	10.46
Crude protein (CP), %	16.50	16.50
Calcium, %	3.60	1.80
Total phosphorus, %	0.60	0.69
Available phosphorus, %	0.35	0.35
Methionine, %	0.40	0.40
Lysine, %	0.85	0.85
Methionine +cystine, %	0.68	0.69

¹Premix provided the following per kg of diet: VA 7500 IU, VE 20.00 IU, VK₃ 2.50 mg, VB₁ 3.45 mg, VB₂ 6.00 mg, VB₆ 2.50 mg, VB₁₂ 0.02 mg, Choline chloride 600 mg, Pantothenic acid 20.58 mg, Folic acid 1.00 mg, Biotin 0.20 mg, Nicotinic acid 27.13 mg, Fe (FeSO₄·H₂O) 75.00 mg, Cu (CuSO₄·5H₂O) 10 mg, Mn (MnSO₄·H₂O) 50.00 mg, Zn (ZnSO₄·H₂O) 43.75 mg, Se (Na₂SeO₃) 0.30 mg, I (KI) 0.44 mg.

²Nutrients levels were calculated values based on the feed formulation.

AME: Apparent metabolism energy; NC: Normal calcium; LC: Low calcium.

Sample collection

Three eggs per replicate were collected weekly for egg quality analysis. At the end of the trial, 6 healthy ducks were randomly selected from each replicate for plasma sampling, body weight measurement, and subsequent euthanasia following approved procedures. Segments (approximately 1 cm) of the duodenum, jejunum, and ileum were excised and fixed in formalin for subsequent intestinal morphology analysis. Cecal contents were collected into sterile enzyme-free tubes, immediately flash-frozen in liquid nitrogen, and stored at −80°C until further analysis. Tibias, femurs, humerus, radius, and ulnas from both sides were dissected and cleaned of adherent muscle and fascia. Left bones were stored at −20°C: half were used for the determination of physical properties, bone mineral density (BMD), and composition, while the other half were used for breaking strength analysis. Right bones were stored at −80°C for gene expression analysis.

Productive performance and egg quality

Daily egg number, total egg weight, and feed consumption were recorded per replicate and averaged over the entire experimental period. Egg weight, yolk weight, and eggshell weight were measured using an electronic balance (PWN124ZH/E, Mingbo Environmental Technology Co., Ltd., Qingdao, China). Eggshell, albumen, and yolk proportions were calculated as the percentage of each component relative to total egg weight. Eggshell thickness and breaking strength were measured using an Egg Shell Thickness Gauge and an Egg Force Reader (Orka Food Technology Ltd., Ramat Hasharon, Israel), respectively. Albumen height and Haugh unit were determined with an Egg Analyzer (Orka Food Technology Ltd.).

Plasma bone metabolic markers

Plasma calcium (Ca) and phosphorus (P) concentrations, as well as alkaline phosphatase (ALP) activity, were assayed using commercial kits (Nanjing Jiancheng Bioengineering Institute Co., Ltd., Nanjing, China). Avian-specific enzyme-linked immunosorbent assay (ELISA) kits (Shanghai Enzyme Biotechnology Co., Ltd., Shanghai, China) were used to quantify the following parameters: 1,25-dihydroxyvitamin D₃ (1,25(OH)₂D₃), calcitonin (CT), parathormone (PTH), bone gla protein (BGP), procollagen type I carboxyl-terminal propeptide (PICP), tartrate-resistant acid phosphatase (TRAP), estrogen (E₂), transforming growth factor-β (TGF-β), fibroblast growth factor 23 (FGF-23), insulin-like growth factor 1 (IGF-1), interleukin-1 (IL-1), interleukin-6 (IL-6), and tumor necrosis factor-α (TNF-α).

Bone physical/mechanical properties, and composition

Fresh bone weight was measured using an electronic analytical balance. Bone length was determined with a vernier caliper, and midshaft circumference was measured using a cotton thread. Subsequently, bone mineral density (BMD) and bone mineral content (BMC) were assessed via a dual-energy X-ray absorptiometry system (DTX-200, Osteometer MediTech, Hawthorne, CA, USA). Bone breaking strength was measured using a texture analyzer (TMS-Pro, Food Technology Corporation, Sterling, VA, USA) equipped with an interchangeable load cell (model ILC-S, force range: 0–1000 N).

Bones were heated at 100°C for 6 min to remove residual surface fascia, then dried at 105°C for 24 h. Samples were degreased in 99.5% ether for 96 h, re-dried at 105°C for 4 h, and weighed. After crushing with a disintegrator, calcium (Ca) and phosphorus (P) contents were determined following Chinese National Standards GB/T 6436-2018 and GB/T 6437-2018, respectively.

Gene expression in the tibia and femur

Aseptically, a sample from the distal end of each bone was excised using bone-cutting forceps, placed in a pre-chilled (liquid nitrogen) mortar, and homogenized. Total RNA was extracted using the TRIzol reagent method. Quantitative real-time polymerase chain reaction (qRT-PCR) was performed on a Bio-Rad CFX96 system (Hercules, CA, USA) to quantify the expression of bone metabolism-related genes. Primers (Table S2) were synthesized by Shenggong Bioengineering Co., Ltd. (Shanghai, China). Relative gene expression levels were calculated using the 2^-ΔΔCt method (Livak and Schmittgen, 2001), with β-actin as the reference gene.

Morphology of duodenum, jejunum and ileum

Intestinal segments were embedded in Tissue-Tek for cryosection preparation (5 μm thickness). Sections were stained with hematoxylin-eosin (H&E) and visualized using a light microscope (Opelco, Washington, DC, USA). Villus height (from the villus tip to the villus-crypt junction) and crypt depth (from the junction to the crypt base) were measured, and the villus height-to-crypt depth (V/C) ratio was calculated. For each intestinal segment, six well-oriented, intact longitudinal villi with adjacent crypts were randomly selected per section per bird for morphometric analysis.

Cecal microbiota profiling via metagenomics analysis

Total genomic DNA was extracted from cecal microbiota using the OMEGA Mag-Bind Soil DNA Kit (M5635-02, Omega Bio-Tek, Norcross, GA, USA). The extracted DNA was used for metagenomic shotgun sequencing library construction with an insert size of approximately 400 bp, using the Illumina TruSeq Nano DNA LT Library Preparation Kit (Illumina, San Diego, CA, USA). Libraries were sequenced on the Illumina NovaSeq platform (Illumina) using the PE150 sequencing strategy at Personal Biotechnology Co., Ltd. (Shanghai, China). Raw sequencing reads were quality-filtered using fastp (v0.23.2) with the following parameters: minimum read length of 50 bp, Phred quality score threshold of Q20, and maximum unqualified base proportion of 30%. Ribosomal RNA sequences were removed using BBDuk (BBMap v38.98; `k = 23, minkmerhits=6`). Taxonomic classification of quality-filtered reads was performed using Kraken2 (v2.0.8-beta) against the NCBI-nt database (accessed 2022-11-01), with abundance re-estimation performed by Bracken (v2.8). All bioinformatics analyses were performed on the GenesCloud platform (Personal Biotechnology Co., Ltd., Shanghai, China; https://www.genescloud.cn).

Cecal microbial metabolite profiling via LC-MS/MS analysis

Cecal contents were homogenized by low-temperature ultrasonic treatment for 30 min, incubated at −20°C for 10 min to precipitate proteins, and centrifuged at 14,000 × g, 4°C for 20 min. The resulting supernatants were stored at −80°C until LC-MS/MS analysis. Metabolite separation was achieved using an Agilent 1290 Infinity UHPLC system (Agilent Technologies) equipped with both HILIC and RPLC columns. Detection was carried out using a quadrupole time-of-flight mass spectrometer (AB SCIEX TripleTOF 6600) coupled to the UHPLC system. Raw MS data (WIFF.Scan files) were converted to mzXML format using ProteoWizard (v3.0.8789) and processed with XCMS (v3.12.0) for peak detection, retention time correction, and alignment. Metabolite identification was performed by matching MS2 spectra against a combination of public databases, including the Human Metabolome Database (HMDB), METLIN, MassBank, and mzCloud, as well as a locally constructed spectral library. Matching criteria included retention time, precursor ion mass (mass tolerance < 10 ppm), MS/MS fragmentation spectra, and collision energy. Metabolite identifications were accepted at confidence Level 2 or above according to the Metabolomics Standards Initiative (MSI) criteria.

Statistical analysis

Each replicate was taken as the experimental unit. The normality of the data and homogeneity of variances were first verified by the Explore procedure using SPSS 16.0 for Windows (SPSS Inc., Chicago, IL). Then the data between groups was conducted using Student’s t-test in SPSS 16.0. These results are expressed as the means and standard errors, and differences were considered to be significant at P < 0.05.

For high-throughput sequencing data, differential abundance analysis of microbial taxa was performed using the GenesCloud cloud analysis platform (Personal Biotechnology Co., Ltd., Shanghai, China; https://www.genescloud.cn). Relative abundance data at the species level were log10-transformed prior to statistical analysis. Two-group comparisons between LC and NC were conducted using Wilcoxon rank-sum -test. To control for the false discovery rate (FDR) arising from multiple testing, p-values were adjusted using the Benjamini-Hochberg (BH) method. Taxa with an adjusted P-value < 0.05 were considered significantly deferentially abundant. Furthermore, linear discriminant analysis (LDA) effect size (LEfSe) was performed to identify significantly differential functional pathways between the two groups, with an LDA score threshold of > 2.0.

Results

Productive performance and egg quality

As shown in Table 2, compared with ducks fed the NC diet, those fed the LC diet exhibited decreased egg production and egg mass, as well as increased feed conversion ratio (FCR) (P < 0.05). Regarding egg quality, the LC diet increased egg shape index and yolk ratio (P < 0.05), while reducing eggshell breaking strength, eggshell thickness, Haugh unit, albumen ratio, and eggshell ratio (P < 0.05). Additionally, albumen height tended to decrease in the LC group (P = 0.052).

Table 2.

Dietary Low-calcium level decreased productive performance and egg quality in laying ducks.

Variables	Dietary Ca level		P-value3
	3.6% (NC)	1.8% (LC)
Productive performance1
Egg production (%)	77.0 ± 3.15	56.0 ± 5.95	< 0.001
Average egg weight (g)	62.8 ± 2.24	63.4 ± 2.89	0.707
Egg mass (g/d)	48.2 ± 2.02	35.4 ± 2.86	< 0.001
Average daily feed intake (g/d/bird)	145.1 ± 8.00	146.5 ± 6.65	0.758
FCR (g:g)	3.09±0.11	4.20±0.36	< 0.001
Cracked eggs (%)	4.36±1.87	4.40±1.78	0.967
Egg quality2
Weight (g)	63.9 ± 1.11	64.5 ± 1.17	0.357
Size shape	1.37±0.01	1.40±0.02	0.046
Breaking strength (N)	39.3 ± 2.19	34.3 ± 1.29	0.001
Thickness (mm)	0.311±0.01	0.282±0.01	0.001
Albumen height (mm)	7.70±0.22	7.40±0.27	0.052
Haugh unit	87.4 ± 1.24	85.3 ± 1.32	0.018
Yolk ratio (%)	31.6 ± 1.02	33.2 ± 0.58	0.007
Albumen ratio (%)	59.1 ± 1.02	57.8 ± 0.29	0.019
Eggshell ratio (%)	9.32±0.12	8.92±0.24	0.004

¹Mean of 6 replicates (12 ducks per replicate) per treatment.

²Mean of 6 replicates (average of 3 eggs per replicate at the end of wk 1, 2, 3, 4 and 5) per treatment.

³It indicates significant differences that is P < 0.05. In the same row, values with different letter superscripts mean significant difference P < 0.05. The same as below.

FCR: feed conversion ratio; NC: Normal calcium; LC: Low calcium.

Plasma bone metabolic markers

Compared with the NC group, ducks fed the LC diet had increased plasma concentrations of 1,25(OH)₂VD₃, PICP, E₂, IGF-1 and IL-1(Table 3, P < 0.05). Plasma PTH, TGF-β and FGF-23 contents tended to increase in the LC group (P < 0.1).

Table 3.

Low-calcium diet affected bone metabolic markers in plasma in laying ducks.

Variables1	Dietary Ca level		P-value
	3.6% (NC)	1.8% (LC)
Ca (mmol/L)	251±47.5	243±36.6	0.760
P (mol/L)	0.283±0.0321	0.256±0.0250	0.136
1,25(OH)2VD3 (ng/mL)	2.81±0.362	3.36±0.186	0.008
CT (pg/mL)	2.60±0.132	2.69±0.214	0.411
PTH (pg/mL)	8.64±0.836	9.45±0.666	0.093
BGP (ng/mL)	12.1 ± 1.318	12.8 ± 1.548	0.430
ALP (ng/mL)	20.8 ± 2.430	20.9 ± 1.657	0.911
PICP (ng/mL)	0.848±0.0660	0.928±0.1019	0.034
TRAP (pg/mL)	134±12.6	126±14.9	0.370
E2 (pg/mL)	186±14.4	233±7.9	< 0.001
TGF-β (pg/mL)	116±13.6	121.45±11.7	0.065
FGF-23 (pg/mL)	35.6 ± 1.39	38.3 ± 6.254	0.067
IGF-1 (ng/mL)	13.0 ± 1.02	14.4 ± 2.93	0.016
IL-1 (pg/mL)	13.3 ± 1.52	15.4 ± 0.70	0.034
IL-6 (pg/mL)	2.95±0.34	2.98±0.23	0.846
TNF-α (pg/mL)	4.61±0.54	4.78±0.47	0.590

¹Mean of 6 replicates (6 ducks per replicate) per treatment.

Ca: calcium; P: phosphorus; 1,25(OH)₂D₃: 1,25-dihydroxyvitamin D₃; CT: calcitonin; PTH: parathormone; OCN: osteocalcin; ALP: alkaline phosphatase; PICP: procollagentypeⅠC-terminal propeptide; TRAP: tartrate-resistant acid phosphatase; E₂: estrogen; TGF-β: transforming growth factor-β; FGF-23: fibroblast growth factor 23; IGF-1: insulin-like growth factor 1; IL-1: interleukin-1; IL-6: interleukin-6; TNF-α: tumor necrosis factor -α. NC: Normal calcium; LC: Low calcium.

Bone physical/ mechanical properties and composition

Table 4 presents the effects of dietary low Ca on bone physical properties. The LC diet decreased fresh weight and fat-free dry weight of the tibia, femur, humerus, and radius (P < 0.05). It also reduced the length of the tibia, radius, and ulna, as well as the midshaft circumference of the radius and fresh weight of the ulna (P < 0.05). Regarding bone mechanical properties and composition (Table 5), the LC diet decreased breaking strength, BMC, BMD, and ash, Ca, and phosphorus (P) contents in the tibia, femur, humerus, radius (except BMC), and ulna (except BMC, BMD, and P content) compared with the NC group (P < 0.05).

Table 4.

Low-calcium diet affected bone physical properties in laying ducks.

Variables1	Dietary Ca level		P-value
	3.6% (NC)	1.8% (LC)
Tibia
Fresh weight (g)	4.52±0.252	4.17±0.156	0.014
Length (mm)	98.4 ± 1.61	96.1 ± 1.32	0.023
Midpoint circumference (mm)	17.6 ± 0.58	17.6 ± 0.66	0.948
Fat-free dry weight (g)	2.93±0.325	2.33±0.216	0.045
Femur
Fresh weight (g)	3.72±0.254	3.32±0.242	0.020
Length (mm)	55.8 ± 1.82	56.0 ± 1.27	0.778
Midpoint circumference (mm)	19.1 ± 0.54	19.0 ± 0.71	0.991
Fat-free dry weight (g)	2.50±0.428	1.44±0.238	< 0.001
Humerus
Fresh weight (g)	5.66±0.466	4.47±0.136	< 0.001
Length (mm)	94.3 ± 2.39	92.7 ± 2.17	0.255
Midpoint circumference (mm)	21.6 ± 0.87	21.5 ± 0.87	0.873
Fat-free dry weight (g)	3.58±0.532	2.34±0.095	0.002
Radius
Fresh weight (g)	2.65±0.240	2.31±0.170	0.017
Length (mm)	79.7 ± 2.55	76.5 ± 1.48	0.022
Midpoint circumference (mm)	17.9 ± 0.59	17.1 ± 0.37	0.022
Fat-free dry weight (g)	1.55±0.230	1.23±0.124	0.014
Ulna
Fresh weight (g)	0.86±0.064	0.77±0.053	0.023
Length (mm)	74.2 ± 0.66	72.5 ± 1.47	0.031
Midpoint circumference (mm)	11.0 ± 0.56	10.7 ± 0.42	0.284
Fat-free dry weight (g)	0.52±0.064	0.48±0.068	0.278

¹Mean of 6 replicates (6 ducks per replicate) per treatment.

NC: Normal calcium; LC: Low calcium.

Table 5.

Low-calcium diet affected bone mechanical properties and composition in laying ducks.

Variables1	Dietary Ca level		P-value
	3.6% (NC)	1.8% (LC)
Tibia
Breaking strength (N)	174±12.9	141±14.6	0.002
Bone mineral content (g)	2.67±0.566	1.28±0.267	< 0.001
Bone mineral density (g/cm2)	0.41±0.068	0.20±0.040	< 0.001
Ash content (%)	68.0 ± 1.66	62.3 ± 2.49	0.001
Ca content (%)	20.9 ± 0.18	18.5 ± 1.39	0.002
P content (%)	11.4 ± 0.16	10.7 ± 0.39	0.002
Femur
Breaking strength (N)	187±21.5	124±10.1	< 0.001
Bone mineral content (g)	1.67±0.182	1.25±0.182	0.003
Bone mineral density (g/cm2)	0.37±0.052	0.28±0.059	0.014
Ash content (%)	69.4 ± 1.42	61.6 ± 5.28	0.006
Ca content (%)	20.6 ± 1.21	17.7 ± 0.35	< 0.001
P content (%)	11.6 ± 0.32	10.5 ± 0.43	0.001
Humerus
Breaking strength (N)	151±11.2	103±13.6	< 0.001
Bone mineral content (g)	1.66±0.240	0.88±0.262	< 0.001
Bone mineral density (g/cm2)	0.36±0.059	0.19±0.070	0.001
Ash content (%)	65.8 ± 1.72	60.7 ± 1.79	< 0.001
Ca content (%)	19.9 ± 0.46	18.5 ± 0.87	0.005
P content (%)	11.1 ± 0.35	10.5 ± 0.19	0.005
Radius
Breaking strength (N)	86.2 ± 12.7	62.3 ± 5.8	0.002
Bone mineral content (g)	0.83±0.070	0.56±0.083	< 0.001
Bone mineral density (g/cm2)	0.08±0.026	0.07±0.011	0.228
Ash content (%)	64.2 ± 01.29	57.5 ± 1.56	< 0.001
Ca content (%)	19.4 ± 0.63	17.8 ± 0.59	0.001
P content (%)	11.0 ± 0.36	10.4 ± 0.25	0.008
Ulna
Breaking strength (N)	48.1 ± 3.6	36.2 ± 5.1	0.001
Bone mineral content (g)	0.20±0.069	0.14±0.041	0.103
Bone mineral density (g/cm2)	0.18±0.077	0.12±0.033	0.122
Ash content (%)	63.2 ± 1.93	60.3 ± 1.13	0.009
Ca content (%)	18.4 ± 1.45	16.7 ± 0.94	0.036
P content (%)	10.5 ± 0.35	10.1 ± 0.27	0.057

¹Mean of 6 replicates (6 ducks per replicate) per treatment. NC: Normal calcium; LC: Low calcium.

Gene expression in tibia and femur

In the tibia, dietary low Ca levels altered the mRNA expression of RANK, TRAP, MMP, RANKL, OPG, ALP and Runx2 compared with the NC group (P < 0.05, Table 6). In the femur, the LC diet upregulated the mRNA expression of RANK, MMP, OPN, OPG, ALP and Runx2 in femur relative to the NC group (P < 0.05).

Table 6.

LC diet affected gene expression in tibia and femur in laying ducks.

Variables1	Dietary Ca level		P-value
	3.6% (NC)	1.8% (LC)
Tibia
RANK	1.74±0.566	4.55±1.105	< 0.001
TRAP	1.08±0.252	6.50±1.648	< 0.001
MMP9	1.27±0.343	2.10±0.657	0.021
OPN	1.54±0.816	1.96±1.219	0.498
RANKL	0.789±0.2963	1.81±0.202	< 0.001
OPG	1.09±0.414	1.68±0.485	0.045
ALP	1.09±0.296	5.48±0.905	< 0.001
Runx2	1.12±0.368	0.165±0.1273	< 0.001
Femur
RANK	0.261±0.0700	0.862±0.2077	< 0.001
TRAP	1.07±0.555	1.59±0.752	0.202
MMP9	1.18±0.488	2.74±0.730	0.001
OPN	1.03±0.655	4.51±1.528	< 0.001
RANKL	1.15±0.481	1.71±0.451	0.062
OPG	0.81±0.291	1.50±0.597	0.030
ALP	1.06±0.414	6.25±0.851	< 0.001
Runx2	1.09±0.528	2.83±1.231	0.010

¹The mean of 6 replicates (6 ducks per replicate) per treatment.

RANK: receptor activator of NF-κB; TRAP: tartrate-resistant acid phosphatase; MMP9: matrix metalloproteinase-9; OPN: osteopontin; RANKL: receptor activator of NF-κB ligand; OPG: osteoprotegerin; ALP: alkaline phosphatase; Runx2: runt-related transcription factor 2.

NC: Normal calcium; LC: Low calcium.

Morphology of duodenum, jejunum and ileum

Compared with the NC group, the villus height of duodenum, the villus height and width of jejunum, the villus height-to-crypt depth (V/C) ratio of duodenum, jejunum and ileum were decreased in laying ducks fed with low Ca level in diet (Fig. 1, P < 0.05).

Fig. 1.

Effects of a LC diet on intestinal morphology of laying ducks.(A) Histological sections of the duodenum, jejunum, and ileum (H&E staining, 5 ×); (B) Villus height, villus width, and crypt depth of the jejunum; (C) Villus height, villus width, and crypt depth of the ileum; (D) Villus height, villus width, and crypt depth of the duodenum; (E) Villus height-to-crypt depth ratio of the jejunum, ileum, and duodenum.

NC: Normal calcium; LC: Low calcium.

Cecal microflora profiling via metagenomics analysis

At the genus level, the cecal microbiota of the NC and LC groups were predominantly composed of Phocaeicola, Gemmiger, Faecalibacterium, Bacteroides, Thermophilibacter, Prevotella, Paraprevotella, Desulfovibrio, Mediterraneibacter and Brachyspira (Fig. 2A). Alpha diversity including the Chao1, Shannon, Simpson and ACE indices, observed species, and Good’s coverage, was evaluated to reflect microbial diversity within individual samples. Compared with the control, the LC diet reduced the Chao1 index, Shannon index, observed species, and ACE index, while increasing Good’s coverage in laying ducks (Fig. 2C, P < 0.05). Beta diversity reflects intersample differences in species composition, visualized via principal coordinate analysis (PCoA). As shown in Fig. 2D, PCoA revealed distinguishable intergroup sample distances, yet the microbial communities of the two groups were similar, with no significant difference between LC and NC groups (PERMANOVA: R² = 0.063, P = 0.866). Differential abundance analysis revealed that the LC diet significantly reduced the relative abundance of Firmicutes (phylum) and Bacilli (class) compared with the NC group (Fig. 2B, P < 0.05). Cecal microbial function analysis indicated overall functional similarity between the two groups (Fig. 2E). Furthermore, LEfSe analysis based on KEGG functional profiles showed that, the LC diet upregulated protein export and selenocompound metabolism, while downregulating xenobiotic biodegradation and metabolism (Fig. 2F, P < 0.05).

Fig. 2.

Cecal microflora profiling via metagenomics analysis. (A) Relative abundance at genus level; (B) Differentially abundant taxa between NC and LC groups; (C) Alpha diversity indices; (D) Functional beta diversity based on predicted KEGG pathways (Bray–Curtis, PCoA); (E) Taxonomic beta diversity (weighted UniFrac, PCoA); (F) Functional pathways significantly enriched in the LC group (LEfSe, LDA > 2).

NC: Normal calcium; LC: Low calcium.

Cecal microbial metabolite profiling via LC-MS/MS analysis

In this study, cecal microbial metabolite between the NC and LC groups was analyzed, and a total of 11 metabolites classes were identified. These included Carboxylic acids and derivatives (20.4%), Fatty acyls (12.4%), Prenol lipids (8.6%), Organooxygen compounds (8.1%), Steroids and steroid derivatives (5.9%), Benzene and substituted derivatives (5.4%), Organonitrogen compounds (4.8%), Azacyclic compounds (3.2%), Organic oxides (3.2%), Phenols (2.7%) and others (25.3%) (Fig. 3A). PCoA analysis revealed significant differences in microbial metabolite profiles between the two groups (Fig. 3B), and 10 differential metabolites classes were subsequently screened for bioinformatics analysis (Fig. 3C). Compared with the NC, the LC diet increased the contents of (S)-piperidine-2-carboxamide, O-acetylcarnitine, confertifolin, ethyl oleate, 5-dehydroavenasterol, lanosterin, ursolic acid, and galactosylsphingosine, while decreasing the contents of picolinic acid and 4-acetamido-2-aminobutanic acid in laying ducks (Fig. 3D, P < 0.05).

Fig. 3.

Cecal microbial metabolite profiling via LC-MS/MS analysis.(A) Metabolite classification. (B) PLS-DA score plot. (C) Screening of differential metabolites (volcano plot). (D) Hierarchical clustering heatmap of differential metabolites.

NC: Normal calcium; LC: Low calcium.

Correlation analysis between cecal microbiota and metabolites

Spearman’s correlation analysis was conducted to explore the relationships between significantly altered microbial taxa and deferentially expressed cecal metabolites (Fig. 4). The results revealed no significant correlations between microbial composition shifts and metabolic variations induced by low-calcium diets, with all pairwise comparisons showing P > 0.05.

Fig. 4.

Spearman correlation heatmap between differentially abundant microbial taxa and differentially expressed cecal metabolites. Heatmap displaying Spearman correlation coefficients (r) between key differential microbial species (rows) and differential metabolites (columns). Red indicates positive correlation; blue indicates negative correlation.

NC: Normal calcium; LC: Low calcium.

Discussion

Calcium plays an essential role in the poultry industry, with extensive research focusing on its function in maintaining eggshell and bone quality in laying hens (Fu et al., 2024; Zhao et al., 2020; Sharma et al., 2023) and ducks (Wang et al., 2014). In the current study, a LC diet decreased egg production and mass, and increased feed conversion ratio (FCR) in laying ducks. These results align with previous findings in laying ducks and hens (Zhao et al., 2020; Xia et al., 2019). The negative impact of low calcium on egg production is primarily attributed to inhibited follicle development. Specifically, Chen et al. (2020) reported that dietary calcium deficiency suppresses follicle selection in laying ducks via a mechanism involving the cyclic adenosine monophosphate (cAMP)-mediated signaling pathway. Notably, calcium transduction is critical for regulating the fate of senescent cells (Martin et al., 2023), and calcium signaling pathways can modulate follicular atresia in aging laying chickens (Yao et al., 2020). In this respect, the lack of calcium in dietary could depressed the follicular development and reduced egg production in laying ducks.

Furthermore, the current study demonstrated that dietary calcium deficiency impairs eggshell quality, as evidenced by reduced breaking strength, thickness and shell ratio. Consistent with extensive literature, calcium plays a pivotal role in eggshell formation (Jin et al., 2024), with its dietary level and source closely associated with shell quality (Xia et al., 2019; Cheng et al., 2025). Specifically, dietary calcium insufficiency compromises shell quality by suppressing shell biomineralization (Chen et al., 2015b). In addition, the Haugh unit, albumen height and albumen ratio of duck eggs were decreased with reduced calcium level in diet. This suggests that the albumen formation was possibly affected by dietary calcium, which was partly due to the changes of structure and quality of the egg white. It was reported that ovomucin (approximately 3.5%) is responsible for the thick gel characteristics of liquid egg whites, and its content of the egg affected the Haugh unit and albumen height (Omana et al., 2010). Besides, the extraction of amino acids from the blood by the magnum corresponds closely with the rate of albumen synthesis (Edwards et al., 1976). In this way, the absorption of protein or amino acids in intestine of laying ducks maybe disturbed with dietary reduced calcium, as founded that the intestinal morphology was damaged in this study. But the precise mechanism of decreased weight and quality of egg white with low dietary calcium level requires further study.

Serum bone remodeling parameters reveal bone metabolism associated with bone fractures in laying hens (Wei et al., 2023). In this study, dietary calcium deficiency increased the contents of 1,25(OH)₂VD₃, PICP, E₂, IGF-1 and IL-1, and tended to increase the PTH, TGF-β and FGF-23 contents in plasma, which inferred that the bone remodeling was activated in laying ducks. The 1,25(OH)₂D₃ can promote absorption of Ca and P in the duodenum to meet the needs of the body (Kraidith et al., 2016). PICP is responsible for forming bone organic matter, in turn related to osteogenesis (Song et al., 2018). E₂ has an effect on bone turnover (Xie et al., 2011), and its deficiency led to increased bone turnover and increased osteoclastic bone resorption owing to enhanced production of a number of proinflammatory and osteoclastogenic cytokines (Pacifici, 1996). IGF-1 functions as a key anabolic regulator of bone cell activity by decreasing collagen degradation and increasing bone matrix deposition and osteoblastic cell recruitment (Yakar et al., 2002), and it declined with the progressive decrease in bone formation markers (Adami et al., 2010). The cytokines affect osteoclast formation and activity to modulate the bone-resorbing process, such as the IL-1, IL-6 and TNF can stimulate the formation and bone-resorbing capacity of osteoclasts, whereas, IL-4 and TGF-β inhibit both osteoclast formation and osteoclast activity (Roodman, 1993). The extracellular calcium could modulate inflammation-mediated (such as IL-1) resorption by means of the parathyroid calcium-sensing receptor (Klein, 2018). PTH induces bone resorption by osteoclasts, and by promoting Ca and inhibiting P reabsorption in kidney (Kuzma et al., 2021). FGF23 has biphasic effects on osteoclast physiology, inhibiting osteoclast formation while stimulating slightly osteoclast activity (Allard et al., 2015). These bone metabolic makers changes mentioned above were probably to modulate the Ca and P metabolism to meet the requirements of production in laying ducks.

Although the bone remodeling of ducks was activated with low calcium level in diet, the bone physical and mechanical properties, and the components were reduced in bones including the tibia, femur, humerus, radius and ulna. These changes indicate a disruption in bone homeostasis, resulting in bone mineral loss and structural impairment (Whitehead, 2004). Mineral reserve in the bones of laying birds is critical for eggshell quality and bone integrity (Kerschnitzki et al., 2014). Bone mineral content and density (BMC and BMD) are key indicators of bone density and strength (Rachner et al., 2011). Reductions in these parameters may be a key factor contributing to decreased eggshell quality in laying ducks. In addition, the dynamic balance between bone formation and resorption is regulated by multiple gene products secreted by osteoclasts and osteoblast. As reported, RANK, TRAP and MMP9 functioned actively during bone resorption to induce osteoclast maturation (Khosla, 2001), promote osteoclast capacity (Hayman et al., 1996), and degraded bone mineral matrix (Colnot et al., 2003), respectively. OPG was the decoy receptor secreted by osteoblasts to compete with RANK for binding to RANKL to block osteoclast maturation process (De Leon-Oliva et al., 2023), and the ratio of RANKL/OPG is considered a critical factor for osteoclast maturation and activation, which dictates the extent of bone resorption (Tobeiha et al., 2020). ALP is a marker of osteoblast differentiation and maturation, and mature osteoblasts mediate mineralization of newly formed bone matrix (Vimalraj, 2020). Runx2 is an essential transcription factor for osteoblast differentiation (Komori, 2018). In the current study, the genes expression in tibia and femur were enhanced with dietary Ca deficiency, including the RANK, TRAP (only in tibia), MMP9, OPN (only in femur), RANKL (only in tibia), OPG, and ALP. These findings indicate that dietary calcium deficiency activates bone remodeling in ducks, involving both osteogenic and osteoclastic pathways. Notably, dietary Ca restriction induced bone site-specific transcriptional responses related to remodeling. Specifically, the tibia showed significant upregulation of the osteoclast-/resorption-related markers TRAP and RANKL, accompanied by an increased RANKL/OPG ratio and decreased Runx2 expression. These changes indicate a resorption-dominant remodeling response, suggesting that bone resorption is favored in the tibia. In contrast, the femur showed an opposite pattern of decreased RANKL/OPG ratio, upregulated Runx2 expression, and no significant changes in TRAP and RANKL levels. These contrasting responses between the tibia and femur likely reflect site-specific bone responses to Ca deficiency, given that the tibia bears greater mechanical loading and contains a higher proportion of cortical bone (Rath et al., 2000; Whitehead, 2004). However, existing evidence is limited to the transcriptional level without proteomic and histomorphometric support, so this site-specific pattern remains only a phenotypic observation rather than an established mechanism. Furthermore, these changes likely contribute to the impaired mechanical properties and reduced mineral content of the tibia and femur in laying ducks fed a low-calcium diet. Although osteogenesis predominates in the femur, the systemic calcium and phosphorus deficiency induced by dietary calcium restriction is insufficient to offset bone mineral loss, ultimately leading to impaired bone quality. Moreover, such divergent effects of dietary calcium deficiency on the tibia and femur constitute a novel phenotypic finding in laying ducks, which merits further mechanistic exploration.

The small intestine is the primary organ for nutrient digestion and absorption in poultry. Villus height and crypt depth reflect the quantity and maturation rate of enterocytes, which further influence the intestinal capacity for nutrient absorption and transport (Wang et al., 2025). In the current study, dietary calcium deficiency reduced villus height and villus height-to-crypt depth ratio in the duodenum, jejunum and ileum of laying ducks. This finding suggests that intestinal absorption capacity was impaired, which may lead to the decreased productive performance and egg quality observed in laying ducks. Similarly, Metzler-Zebeli et al. (2012) reported that a low calcium-phosphorus diet increased cecal crypt depth in young pigs, indicating that dietary calcium-phosphorus levels can modulate intestinal tissue responses. In addition, dietary supplementation with active dicalcium phosphate has been shown to promote intestinal development in broiler chickens, it increases villus height and the villus height-to-crypt depth ratio while reducing crypt depth, thereby improving growth performance (Xing R, 2020). In this respect, the impaired intestinal development may lead to insufficient nutrient digestion and absorption, which in turn contributes to reduced productive performance in laying ducks fed a LC diet.

This study used metagenomic sequencing to characterize how a LC diet affects the structural and functional dynamics of cecal microbiota in laying ducks. Our findings show that dietary calcium deficiency induces functional adaptations in the gut microbial community. Previous studies have reported microbiota-specific modulation under low-protein diets, such as increased abundance of Bacteroides (Zhou et al., 2024), proliferation of short-chain fatty acid-producing bacteria (Li et al., 2024), and stabilization of microbial communities (Chang et al., 2023). Consistent with these reported patterns, the LC diet in this study significantly reduced microbial species richness and evenness. It also specifically suppressed beneficial taxa in the phylum Firmicutes, which are involved in carbohydrate fermentation. Furthermore, the microbiota showed extensive metabolic reprogramming. Such reprogramming was manifested as coordinated activation of iron transport, oligopeptide uptake and carbohydrate metabolism pathways, accompanied by strengthened global transcriptional regulation. These findings indicate that LC diet feeding is associated with the reallocation of microbial metabolic resources toward basal nutrient acquisition and endogenous metabolic homeostasis, which may reflect a conserved adaptive strategy of gut microbiota in response to nutritional stress.

In addition, dietary low Ca intake altered cecal metabolite profiles, with 8 metabolites increased and 2 decreased. Among these, ursolic acid (UA) and picolinic acid (PA) were identified as key differential metabolites associated with bone metabolism. UA has been reported to exert anabolic effects on bone by enhancing trabecular microarchitecture, suppressing osteoclast activity, and promoting osteoblast differentiation via the mitogen-activated protein kinase (MAPK), nuclear factor-kappa B (NF-κB), and activator protein-1 (AP-1) signaling pathways (Zheng et al., 2020; Lee et al., 2008)). Although components of the MAPK/NF-κB/AP-1 signaling pathways were not directly quantified in bone tissue in the present study, their downstream targets TRAP and MMP9 were increased under Ca deficiency, which is consistent with observed changes in bone metabolism and quality in the current study. In contrast, PA is a tryptophan-derived metabolite that has been shown to stimulate osteoblast activity and enhance bone formation in vivo (Duque et al., 2020). Specifically, low Ca increased cecal UA levels while reducing PA levels, which may reflect a gut microbial compensatory response to calcium deficiency. A plausible explanation is that Ca deficiency reshapes the cecal microenvironment and microbial metabolic activity, thereby shifting the balance of bone-related metabolites derived from diet–microbiota–host interactions (Jiang et al., 2023). Given that both UA and PA have been reported to exert pro-osteogenic effects, such opposing variations may reflect concurrent shifts in bone-related metabolites that are potentially associated with the gut–bone axis. Furthermore, we also detected a marked increase in galactosylsphingosine in the present study. As a key sphingolipid metabolite, this molecule may modulate calcium-phosphorus homeostasis and osteocyte function by regulating cell membrane composition, cellular signal transduction and oxidative stress pathways (Ahn et al., 2015). Although the correlation analysis between metagenome and metabolome showed no statistical significance, their coordinated changes at the module level still imply potential associations. While these findings suggest links between gut microbiota and other key phenotypic indicators, the definite biological functions of relevant molecules in the gut–bone axis under low-calcium stress remain poorly clarified. Further validations via targeted metabolomics, correlation network analysis and functional experiments are therefore required.

Notably, although clear skeletal and intestinal damage was observed in this study, changes in cecal microbiota composition and metabolite profiles were relatively modest. This apparent discrepancy may be associated with the fact that host tissues including bone and intestinal epithelium are more sensitive to calcium deficiency and respond more rapidly, whereas the gut microbiota exhibits greater resilience and a delayed adaptive response (Fragiadakis et al., 2020). In addition, the 5-week feeding period may be insufficient to reflect the long-term adaptive succession of the gut microbiota and the steady state of bone remodeling homeostasis. Moreover, the functional redundancy of the gut microbiome allows for the preservation of core metabolic functions, even with subtle alterations in taxonomic composition (Moya and Ferrer, 2016). Notably, calcium deficiency-mediated bone loss in the present model is primarily associated with systemic endocrine signaling, including elevated PTH, 1,25(OH)₂D₃, and E2. These factors may directly regulate bone remodeling with minimal reliance on microbiota-derived pathways (Kuzma et al., 2021; Kraidith et al., 2016). These potential associations indicate that gut microbiota and their metabolites may be linked to bone metabolism, yet their precise contribution to this process remains to be fully elucidated. Future studies may perform fecal microbiota transplantation (FMT) and metabolite supplementation assays to elucidate the causal relationship between gut microbiota-derived metabolites and bone metabolism. Furthermore, this study is restricted to a single low-calcium level (1.8%), lacking both dose-gradient settings and positive control group. This hinders the exploration of dose-response relationships as well as potential interactions with other nutrients including phosphorus and vitamin D₃. Future studies with graded Ca levels are required to fully elucidate these dose-dependent effects.

Conclusion

A low-Ca diet can activate bone remodeling by upregulating the expression of bone metabolism-related genes and altering the cecal microbiota-metabolite milieu, but this adaptive response is insufficient to maintain skeletal homeostasis and intestinal structural integrity in aged laying ducks, ultimately leading to impaired bone integrity and decreased eggshell quality. Notably, a low-Ca diet exerts differential regulatory effects on the tibia and femur of laying ducks, suggesting a bone site-specific phenotypic response to calcium deficiency that warrants further mechanistic investigation. This study clarifies the physiological and molecular regulatory mechanisms underlying the response of aged laying ducks to LC nutrition, and provides a scientific basis for improving bone health and productive performance of aged laying ducks through optimized Ca nutritional strategies.

CRediT authorship contribution statement

Yuhong Zhang: Writing – original draft, Project administration, Formal analysis, Conceptualization. Li Min: Writing – original draft, Project administration, Formal analysis, Conceptualization. Chuntian Zheng: Resources, Project administration, Investigation. Wei Chen: Visualization, Validation, Data curation. Shuang Wang: Visualization, Validation, Data curation. Yanan Zhang: Writing – review & editing, Supervision, Methodology.

Disclosures

We declare that we have no financial and personal relationships with other people or organizations that can inappropriately influence our work, and there is no professional or other personal interest of any nature or kind in any product, service and/or company that could be construed as influencing the content of this paper.