Skip to main content
NCBI home page
Food Chemistry: X logoLink to Food Chemistry: X
. 2026 Sep 13;39:104417. doi: 10.1016/j.fochx.2026.104417

Comparative characterization of meat quality and volatile compounds in raw meat from Magang geese (Anser cygnoides) and Hungarian geese (Anser anser) based on LC-MS and HSSPME-GC-MS

Kaiqi Weng a,1, Yi Liu a,1, Huiying Wang a, Yu Tan b, Daqian He a,⁎
PMCID: PMC13594591  PMID: 42774433

Abstract

Genetic background is a key determinant of meat quality and flavor-related chemical characteristics, yet the metabolic basis of breed-dependent differences in goose meat remains unclear. Here, Magang (MG, Anser cygnoides) and Hungarian (HB, Anser anser) geese were compared using muscle histology, physicochemical analysis, and integrated LC-MS and HSSPME-GC-MS profiling of raw breast muscle. HB geese showed higher intramuscular fat and lower pressing loss, whereas MG geese exhibited higher protein and moisture contents. Metabolomic analysis revealed pronounced breed-specific differences, with MG enriched in amino acid- and nucleotide-related metabolites, while HB displayed higher levels of lipid-related metabolites, particularly phospholipids. Volatile analysis further revealed higher abundances of esters and alcohols and reduced lipid oxidation-derived ketones in MG meat, indicating breed-related differences in volatile compound profiles. Overall, these results demonstrate that breed-dependent differences in amino acid and lipid metabolism are associated with variations in nutritional composition and volatile compound profiles in goose meat.

Keywords: Goose meat, Metabolomics, Volatile compounds, Meat quality

Highlights

  • •

    Hungarian geese exhibit higher intramuscular fat and water-holding capacity, while Magang geese show higher protein content.

  • •

    Breed-specific regulation of amino acid, purine, and lipid metabolism shapes nutritional composition.

  • •

    Distinct volatile profiles indicate reduced lipid oxidation and enhanced ester-related flavor in Magang geese.

1. Introduction

Meat quality not only includes appearance and eating quality traits, but also encompasses nutritional value and sensory attributes, thereby directly influencing consumer acceptance and health perceptions of meat products (Joo et al., 2013). Goose is the third largest poultry species in terms of annual slaughter volume, with production in China reaching 5.10 million tons (FAO-STAT, 2023). Owing to its large production scale, goose meat has attracted increasing attention for its distinctive fatty acid profile, particularly its high proportion of polyunsaturated fatty acids, which are considered nutritionally advantageous compared with those of red meats (Ismail & Joo, 2017).

Recent studies have shown that both postmortem storage conditions and genetic background significantly influence the metabolic composition and volatile flavor profile of goose meat (Chen et al., 2024; Guo et al., 2024). Consistently, our previous work identified distinct nutritive profiles and potential biomarkers for the authentication of goose meat from older birds (Weng et al., 2022). However, the goose industry in China is highly regionalized, relying predominantly on indigenous local breeds, while certain foreign goose breeds have also been introduced for commercial production or crossbreeding (Krunt & Zita, 2025). Magang geese and Hungarian geese represent typical examples of Chinese indigenous and Western improved goose breeds, respectively. Despite their marked differences in growth performance, meat quality traits, processing suitability, and market demand, systematic comparisons of their nutritional composition and flavor-related metabolic characteristics remain limited, and the underlying biochemical mechanisms are still poorly understood.

Metabolomics has emerged as a powerful analytical strategy for elucidating the biochemical basis of meat quality (Muroya et al., 2020). Non-targeted LC-MS approaches enable comprehensive profiling of non-volatile metabolites associated with nutritional value and freshness, whereas HSSPME-GC-MS allows sensitive characterization of volatile organic compounds responsible for meat aroma and flavor (Fiehn, 2016). Using integrated GC-MS and LC-MS approaches, Chen, Cao, Geng, et al. (2023) revealed pronounced differences in amino acids, fatty acids, and bioactive compounds between chicken breeds, demonstrating the effectiveness of metabolomics in discriminating breed-related meat quality traits. Similarly, Tan et al. (2025) reported that sex-related differences significantly affected lipid composition, amino acid profiles, and volatile flavor compounds in Yanbian yellow cattle, further highlighting the value of combined metabolomic and volatilomic analyses in uncovering mechanisms underlying meat quality variation. Together, these findings highlight the value of metabolomics-based strategies in elucidating meat quality–related metabolic characteristics. However, integrated analyses combining non-volatile metabolite profiling with volatile flavor compound characterization in comparative studies of different goose breeds are still scarce. Although the aroma of raw meat is relatively mild, some volatile compounds present in raw meat may persist after cooking, and the odour characteristics of raw meat may also influence consumer acceptance before purchase (Bueno et al., 2019).

In this study, Magang geese (Anser cygnoides) and Hungarian geese (Anser anser) were compared as representative Chinese indigenous and Western improved breeds to elucidate breed-related differences in goose meat quality. An integrated strategy combining LC-MS-based metabolomics and HSSPME-GC-MS was employed to characterize variations in non-volatile metabolites and volatile compounds in raw breast muscle. This study aims to clarify the biochemical basis underlying breed-dependent differences in nutritional composition and flavor-related chemical profiles of breast muscle, thereby providing a scientific foundation for breeding improvement and product differentiation.

2. Materials and methods

2.1. Animals and sample collection

All animal procedures were performed in accordance with the Animals (Scientific Procedures) Act 1986, EU Directive 2010/63, and associated guidelines, and were approved by the Institutional Animal Care and Use Committee of the Shanghai Academy of Agricultural Sciences (Permit No. SAASXM0524061). Commercial Magang geese (Anser cygnoides) and Hungarian geese (Anser anser) were provided by Hunan Wugang Tong Geese Agricultural Development Co., Ltd., China. The Magang goose (MG) is a prominent indigenous Chinese breed, valued for its fast growth and high-quality meat. It is easily recognized by its distinctive ‘three black’ traits: black head, beak, and flippers. The Hungarian goose (HB), developed by the Hungarian Holdobaki Goose Corporation, is a highly esteemed dual-purpose breed renowned for its exceptional down quality and impressive egg production. Both breeds were reared from 1 day of age on the same commercial farm under the same environmental and management conditions. Feeding was performed according to the national standards of Nutrient Requirements of Geese (NY/T4641-2025), and both breeds were provided the same commercial diet (Table S1). In the present study, 20 male Magang geese and 20 Hungarian geese were slaughtered on the same day at 70 days of age. Each bird was individually stunned, then exsanguinated by severing the jugular vein and carotid artery on one side of the neck within 15 s. Samples of breast muscle (Pectoralis major) and leg muscle (Gastrocnemius) were collected, cooled, and stored at 4 °C. At 24 h post-mortem, samples (1 × 1 × 1 cm3) for haematoxylin and eosin staining (H&E) were fixed in 4% paraformaldehyde while samples for LC-MS (1 g) and HSSPME-GC-MS (1 g) were snap-frozen in liquid nitrogen and stored at −80 °C. The remaining muscle samples were collected for meat quality assessment at 24 h post-mortem. Raw breast muscle was used for LC-MS and HSSPME-GC-MS to characterize intrinsic breed-related metabolite and volatile profiles without introducing additional variation caused by cooking.

2.2. Meat quality traits measurements

Meat quality was evaluated based on physical traits (pH24h, meat color, pressing loss, and shear force) and proximate composition (moisture, intramuscular fat, and protein). Samples were obtained from six randomly selected geese per group, with each parameter measured in triplicate. pH24ₕ was determined at three muscle depths (upper, middle, and lower) at 10 mm beneath the surface using a portable pH meter equipped with a penetration electrode (DELTA 320, Mettler Toledo, Columbus, OH, USA). Meat color (L*, a*, b*) was measured using a Minolta CR400 colorimeter (Konica Minolta, Tokyo, Japan) according to the CIE-LAB system. Shear force was determined using a muscle shear device (RH-N50, Guangzhou Runhu Instrument Co., Ltd., China). Cylindrical cores (1 cm diameter) were extracted from various locations of the breast muscle, aligned with the fiber orientation. Pressing loss was measured using a strain-controlled unconfined compression device (YYW-2 dilatometer). Approximately 1 g of sample (W₁) was placed between 16 layers of filter paper on both sides and compressed between plastic plates using the dilatometer at 68.66 kPa for 5 min. The sample was then reweighed (W₂), and pressing loss was calculated as: Pressing loss (%) = [(W1–W2)/W1] × 100%. For proximate composition analysis, breast muscle samples were trimmed of tendons and membranes, cut into small pieces, and homogenized using a high-speed grinder. Moisture, intramuscular fat, and protein contents (g/100 g) were determined using a near-infrared food analyzer (MultiScan Series 3000, Next Instruments Pty Ltd., Australia), calibrated with manufacturer-provided models widely applied for routine proximate analysis of meat. Instrument performance and prediction accuracy were further verified using our previously obtained reference data from conventional chemical analyses, with an overall prediction accuracy of approximately 96%.

2.3. Muscle fiber morphology

Muscle fiber morphology (including diameter and cross-sectional area) was analyzed following haematoxylin and eosin (H&E) staining. Tissues were fixed in 4% paraformaldehyde for 24 h, paraffin-embedded, and sectioned at a thickness of 5 μm using a microtome (Leica Biosystems, Wetzlar, Germany). After overnight drying at 40 °C, sections underwent H&E staining. Quantitative measurements of fiber morphology were conducted using Image-Pro Plus software (Media Cybernetics, Rockville, MD, USA).

2.4. LC-MS

2.4.1. Sample preparation and metabolite extraction

The breast muscle samples used for LC-MS were obtained from the same set of biological samples used for meat quality measurements, with six biological replicates per group (n = 6). Muscle samples (50 mg) were homogenized twice at 55 Hz for 60 s in 200 μL pre-chilled water with two steel beads. After adding 800 μL methanol:acetonitrile (1:1, v/v), the mixture was sonicated for 30 min and incubated at −20 °C for 30 min. Following centrifugation (12,000 rpm, 4 °C, 10 min), 800 μL of supernatant was collected and dried under vacuum. The residue was dissolved in 150 μL 50% methanol containing 5 ppm 2-chlorophenylalanine, vortexed for 30 s, and centrifuged again. The final supernatant was filtered (0.22 μm) and transferred to vials for analysis. A pooled QC sample was prepared by mixing 10 μL from each sample to assess analytical stability.

2.4.2. LC-MS conditions

Chromatographic separation was carried out on an ACQUITY UPLC HSS T3 column (1.8 μm, 2.1 × 100 mm) at 40 °C with a flow rate of 0.4 mL/min. The autosampler was maintained at 8 °C, and the injection volume was 2 μL. Mass spectrometry was performed on a Thermo Orbitrap Exploris 120 with a HESI source in both positive and negative ion modes, controlled by Xcalibur software (v4.7). Key parameters included a spray voltage of +3.5/−3.0 kV, sheath gas at 40 arb, auxiliary gas at 15 arb, capillary temperature at 325 °C, and auxiliary gas temperature at 300 °C. Full MS scans were acquired at 60,000 resolution (m/z 100–1000), with AGC target set to Standard and Max IT at 100 ms. The top 4 precursor ions were selected for MS/MS (resolution 15,000, HCD 30%, dynamic exclusion 8 s). QC samples were injected 2–4 times prior to analysis for system equilibration, and once every 5–10 samples during acquisition to monitor stability and data quality.

2.4.3. Bioinformatic analysis

Raw data files were processed using Compound Discoverer™ 3.3 (v3.3.2.31, Thermo Fisher Scientific, Waltham, MA, USA). Peak extraction, alignment, and correction were performed using the software's updated algorithms for peak detection and quality scoring. Background noise and low-quality peaks were minimized through quality-based filtering. Peaks missing in more than 50% of QC samples were excluded, and remaining missing values were imputed using the Fill Gaps algorithm. The reproducibility of the analytical platform was evaluated using QC samples. Features with a relative standard deviation (RSD) < 30% in QC samples were considered stable, and at least 70% of features meeting this criterion is generally regarded as acceptable. In this study, 88.1% and 92.34% of the LC-MS features retained after initial peak processing in the positive and negative ion modes, respectively, had QC-RSD values <30% (Fig. S1C and D). These percentages were calculated at the feature level prior to metabolite annotation and therefore were not restricted to confidently identified metabolites. Total peak areas were normalized across samples. Metabolite identification was conducted using the in-house PSNGM database, along with public spectral libraries including mzCloud, LIPID MAPS, HMDB, MoNA, and NIST_2020_MSMS. Metabolite annotation confidence was evaluated according to the Metabolomics Standards Initiative criteria, and the identified metabolites were mainly classified as MSI level 2 based on accurate mass and MS/MS spectral matching. The main parameters for database searching were as follows: MS1 tolerance for identification 0.01 Da, MS2 tolerance for identification 0.05 Da, Smoothing level 3, Minimum peak height 10,000, Minimum peak width 5, Mass slice width 0.05, Identification score cutoff 70.

Metabolite abundance data were analyzed using the Pheatmap package (v1.0.12, R) for bidirectional hierarchical clustering of both samples and metabolites, with heatmaps generated to visualize metabolic profiles (Chong & Xia, 2018). Principal component analysis (PCA), partial least squares discriminant analysis (PLS-DA), and orthogonal PLS-DA (OPLS-DA) were performed using the Ropls package to reduce dimensionality and explore group separation. Differential metabolites were screened based on VIP > 1 and p < 0.05, with fold change (FC) values used to assess effect size.

Further analyses included clustering of differential metabolites using Pheatmap, as well as trend plots, Venn diagrams (VennDiagram v1.7.3), and UpSet plots (UpSetR v1.4.0) for visualizing overlaps between comparisons. Correlation analysis (corrplot v4.0.3), box plots, and violin plots (ggplot2 v3.4.1) were used to explore abundance patterns. Machine learning analysis was performed using mlr3verse (v0.2.7), and ROC curves were generated via pROC (v1.18.2) to identify potential biomarkers. Functional enrichment of differential metabolites was conducted using clusterProfiler (v4.6.0) for KEGG pathway analysis (Kanehisa & Goto, 2000), identifying significantly enriched pathways and calculating overall differential abundance scores to assess pathway-level changes and highlight key metabolic alterations.

2.5. HSSPME-GC-MS

2.5.1. Sample preparation and extraction

The breast muscle samples used for HSSPME-GC-MS were derived from the same biological set, with six biological replicates per group (n = 6). Muscle tissue samples were ground into powder in liquid nitrogen. Subsequently, 0.2 g of the powdered sample was accurately weighed and transferred into a 20 mL headspace vial (Agilent Technologies, Palo Alto, CA, USA) preloaded with 0.2 g of sodium chloride (NaCl), which served to inhibit residual enzymatic activity (Longo et al., 2019). The vial was immediately sealed using a crimp-top cap fitted with a polytetrafluoroethylene (PTFE)-lined silicone septum (Agilent Technologies). To extract volatile compounds, headspace solid-phase microextraction (HS-SPME) was employed. The sealed vial was pre-incubated at 60 °C for 5 min to allow equilibration. A fiber coated with divinylbenzene/carboxen/polydimethylsiloxane (DVB/CAR/PDMS; 120 μm thickness, Agilent Technologies) was then inserted into the vial headspace and exposed to the volatiles for 15 min at the same temperature.

2.5.2. GC-MS conditions

After extraction, the volatile organic compounds (VOCs) adsorbed on the solid-phase microextraction fiber were thermally desorbed in the injection port of a gas chromatograph (GC, Model 8890; Agilent Technologies, USA). The desorption was conducted at 250 °C for 5 min in splitless mode to ensure complete transfer of analytes. An Agilent 8890 gas chromatograph coupled with a 7000D triple quadrupole mass spectrometer was employed for compound separation and detection. A DB-5MS capillary column (30 m × 0.25 mm inner diameter, 0.25 μm film thickness; 5% phenyl-methylpolysiloxane) was used for chromatographic separation. Helium, of high purity, served as the carrier gas at a flow rate of 1.2 mL/min. The injection temperature was maintained at 250 °C.

The GC oven temperature was initially set at 40 °C and held for 3.5 min, then increased to 100 °C at a rate of 10 °C/min, followed by a ramp to 180 °C at 7 °C/min, and finally to 280 °C at 25 °C/min, with a 5 min hold at the final temperature. Mass spectrometric detection was performed using electron impact ionization at 70 eV. The quadrupole analyzer, ion source, and transfer line were maintained at 150 °C, 230 °C, and 280 °C, respectively. A widely targeted volatilomics approach was applied using selected ion monitoring (SIM) mode for the targeted acquisition of predefined volatile compounds. A self-established database was constructed based on multiple species, published literature, partial reference standards, and retention indices (RI), including predefined retention times (RT) and characteristic ions for qualitative and quantitative analyses (Shanghai Personal Biotechnology Co., Ltd., Shanghai, China). Each compound was assigned one quantitative ion and two to three qualitative ions, which were monitored in a time-segmented manner according to their elution order. This widely targeted approach was intended for broad screening of potential volatile compounds rather than absolute identification of all detected VOCs. VOC annotation was based on RT, RI, and characteristic-ion matching after background subtraction. Because not all compounds were individually confirmed with authentic standards under the same chromatographic conditions, the detected VOCs are reported as database-based annotations rather than absolute identifications. Quantification was performed using the quantitative ion for peak integration and correction. Detailed analytical procedures were performed as previously described by Yuan et al. (2022).

2.5.3. Bioinformatic analysis

Principal component analysis (PCA) was performed using the prcomp function in R (www.r-project.org) after unit variance scaling of the data. Hierarchical cluster analysis (HCA) and Pearson correlation coefficients (PCC) between samples were visualized as heatmaps using the ComplexHeatmap package in R. HCA included dendrograms for both samples and metabolites, while PCC was shown as correlation heatmaps. Metabolite intensities were unit variance scaled and displayed as color gradients. For two-group comparisons, differential metabolites were identified based on VIP > 1 and |log₂FC| ≥ 1. VIP scores were obtained from OPLS-DA using the MetaboAnalystR package. Model validity was assessed via 200-permutation testing to prevent overfitting. To further explore the relationships between metabolomic and volatilomic profiles, a correlation network was constructed based on differential metabolites and VOCs. Spearman correlation analysis was performed to evaluate cross-platform associations, and only strong and statistically significant correlations (|cor| > 0.8, p < 0.05) were retained.

2.6. Quantitative descriptive sensory analysis

Quantitative descriptive sensory analysis was performed to characterize the aroma of raw goose breast meat according to Katiyo et al. (2020), with minor modifications. Ten panelists (five males and five females, aged 25–50 years) were recruited from the Shanghai Academy of Agricultural Sciences. All panelists were familiar with handling or purchasing raw goose meat. Prior to formal evaluation, the panelists received training sessions to familiarize themselves with the aroma descriptors, standard references, and corresponding intensity scales (Table 1). Five aroma attributes, including bloody, pungent, fishy, rotten-egg, and ammonia-like odors, were evaluated.

Table 1.

Quantitative descriptive sensory analysis of the aroma of raw goose meat.

Descriptor Definition Rating scale Standard reference (extreme level)
Bloody Aromatic associated with raw lean meat, blood, serum or metal/iron 0 = no bloody odour
10 = intense bloody odour		 Raw beef liver
Pungent Very strong, sharp smell 0 = no pungent odour
10 = intense pungent odour		 Reminiscent of a rotten freshwater
fish
Fishy Aromatic associated with spoiled fish 0 = no fishy odour
10 = intense fishy odour		 Reminiscent of a rotten freshwater
fish
Rotten egg Aromatic associated with rotten eggs 0 = no rotten egg odour
10 = intense rotten egg odour		 Reminiscent of a rotten egg
Ammonia-like Aromatic associated with ammonia 0 = no ammonia-like odour
10 = intense ammonia-like odour		 0.2% Ammonia in water
Open in a new tab

During sensory evaluation, raw breast meat samples were labelled with randomly generated three-digit codes. The intensity of each aroma attribute was scored on a 0–10 scale, where 0 indicated no detectable odour and 10 indicated an extremely intense odour. A 60-s rest period was provided between successive samples, during which panelists smelled the back of their hands to minimize olfactory carryover.

2.7. Statistical analysis

Repeat measurements among groups were analyzed using SPSS (version 27.0; SPSS Inc., Chicago, IL, USA) with paired comparisons. Data are presented as mean ± standard error (SE), covering muscle meat quality traits and fiber characteristics. Statistical significance was defined as P ≤ 0.05.

3. Results

3.1. Divergence in meat quality traits and muscle fiber characteristics between the MG and HB geese

Representative muscle fiber characteristics (Fig. 1B) and meat quality traits were determined to reveal the differences in breast and leg muscle between MG and HB geese. The results revealed a significant interaction between breed and muscular tissues on both muscle fiber diameter and cross-sectional area (Fig. 1C). In both goose breeds, the diameter of muscle fiber in the leg muscle was significantly greater, approximately 2.5-fold than that in the breast muscle (P < 0.05). No significant difference was observed in breast muscle size between the MG goose and HB goose. However, the muscle fiber diameter of the leg muscle in MG goose was significantly smaller than that in HB goose (P < 0.05).

Fig. 1.

Fig. 1

Open in a new tab

Difference in muscle fiber characteristics between MG and HB goose breast meat samples. (A) Schematic of MG and HB goose. (B) Haematoxylin and eosin staining, Magnification = 200× (Bar = 100 μm). (C) Values of muscle fiber diameter, cross-sectional area, and fiber density for MG and HB goose. All histograms represent mean ± standard error (n = 6). Statistically significant differences are indicated by different letters (P < 0.05).

To compare the physical properties and proximate composition of meat from different goose breeds, pH24h, meat color, pressing loss, and shear force were measured (Fig. 2). Additionally, muscle moisture content, protein content, and intramuscular fat content were analyzed using a near-infrared food analyzer. Except for shear force and moisture content, significant interactions were observed between breed and muscular tissue for all meat quality traits (P < 0.05). In terms of physical properties, the leg muscle of HB goose exhibited significantly lower pressing loss compared to other groups (P < 0.05). Across all groups, shear force showed a positive trend with muscle fiber diameter, and the shear force of breast muscle was lower than that of leg muscle. Regarding proximate composition, the leg muscle of MG goose had higher protein and moisture contents compared to that of HB goose, whereas HB goose exhibited higher intramuscular fat deposition. In both breeds, the breast muscle had a higher protein content but lower intramuscular fat content than the leg muscle (P < 0.05).

Fig. 2.

Fig. 2

Open in a new tab

Difference in meat quality traits between MG and HB goose breast meat samples. (A) Values of physical properties (pH24h, pressing loss, shear force, and meat color) for MG and HB goose meat. (B) Values of proximate composition (moisture, protein, and IMF content) from MG and HB goose meat. All histograms represent mean ± standard error (n = 6). Statistically significant differences are indicated by different letters (P < 0.05).

Overall, compared with HB geese, MG geese showed higher protein and moisture contents in leg muscles, while HB geese exhibited lower pressing loss and greater intramuscular fat deposition. Additionally, leg muscles had larger fiber diameters and higher shear force than breast muscles in both breeds.

3.2. Metabolomic profiles of goose meat

To identify metabolomic profiles between MG and HB geese, metabolomic sequencing was performed on goose breast meat. Representative TICs of QC samples are shown in Fig. S1, while the corresponding MS/MS data are provided in Table S2, indicating good analytical stability and reliable metabolite identification, as supported by over 88% of features exhibiting RSD < 30% in both ion modes. Pearson correlation coefficient was performed to evaluate correlations between samples (Fig. 3A). A total of 787 metabolites were detected and divided into 10 classes, mainly comprising 186 lipids and lipid-like molecules, 164 organic acids and derivatives, 114 organoheterocyclic compounds, 63 benzenoids, 46 organic oxygen compounds, 36 Nucleosides, nucleotides, and analogues, 30 phenylpropanoids and polyketides, 21 organic nitrogen compounds, 6 alkaloids and derivatives, and others (Fig. 3C). Unsupervised PCA was performed to give a comprehensive view of the clustering trend for the metabolomic profiles. PCA revealed distinct clusters for each group, indicating significant metabolic heterogeneity (Fig. 3B).

Fig. 3.

Fig. 3

Open in a new tab

Metabolomic profiles between MG and HB goose breast meat samples. (A) Pearson correlation coefficients between MG and HB goose breast meat samples. (B) Principal component analysis (PCA) of groups with six biological replicates. (C) Classification of the 787 metabolites of MG and HB goose breast meat samples (n = 6).

Taken together, the metabolomics of goose meat were mainly lipids and lipid-like molecules, organic acids and derivatives, organoheterocyclic compounds, benzenoids, and organic oxygen compounds. Goose meat had limited variation within group and separated well from different breeds.

3.3. Identification and comparison of differential metabolites

To identify differential metabolites between MG and HB goose meat, we performed pairwise comparisons using an OPLS-DA model (Fig. 4A), which showed predictive ability (R2X = 0.22, R2Y = 1.00, Q2 = 0.611). Permutation testing confirmed that the model was not overfitted, as all permuted Q2 values were lower than that of the original model and the Q2 intercept was close to zero (Fig. 4B). Metabolites with VIP > 1 and P < 0.05 were considered significantly different (Table S3). Notably, false discovery rate (FDR) correction using the Benjamini-Hochberg method was also performed, and the adjusted p-values are provided in Table S3. The combined use of VIP and nominal p-values for screening differential metabolites is widely applied in metabolomics studies to balance statistical sensitivity and biological relevance.

Fig. 4.

Fig. 4

Open in a new tab

Identification of differential metabolites between MG and HB goose breast meat samples. (A) OPLS-DA score plots illustrating the separation between MG and HB goose. (B) OPLS-DA permutation test confirming model validity. (C) Volcano plots highlighting significantly altered metabolites. (D) Hierarchical clustering analysis depicting distinct metabolic patterns across samples.

Among the differential metabolites, 94 were down-regulated and 54 were up-regulated in MG compared with HB (Fig. 4C). A substantial proportion of the most strongly down-regulated metabolites were lipids and lipid-like molecules, including LysoPC (P-18:1/0:0), 1-stearoyl-2-arachidonoyl-PS, dihomo-γ-linolenic acid ethyl ester, geranyl phosphate, and PC (P-15:0/0:0). Several metabolites involved in muscle energy production and amino acid metabolism, such as acetyl coenzyme A, pantothenic acid, L-tyrosine, threonyltyrosine, leucylproline, and Ser-Met, also exhibited lower levels in MG relative to HB. In contrast, most of the up-regulated metabolites were organic acids and their derivatives, including balenine, L-arginine, L-proline, Glu-Leu, DL-norvaline, 1-methylproline, and methylmalonic acid. Additionally, metabolites associated with energy metabolism and nucleotide turnover, such as l-carnitine, inosine, uric acid, S-adenosylhomocysteine, and methylthioadenosine, were elevated in MG compared with HB. The clustering heatmap of differential metabolites was shown in Fig. 4D.

Collectively, these results indicated that more metabolites were down-regulated than up-regulated in MG compared with HB. Up-regulated metabolites included organic acids and their derivatives, such as balenine, L-arginine, L-proline, Glu-Leu, DL-norvaline, 1-methylproline, and methylmalonic acid, as well as l-carnitine, inosine, uric acid, S-adenosylhomocysteine, and methylthioadenosine. The down-regulated metabolites were mainly lipids and lipid-like molecules, along with key intermediates of muscle energy and amino acid metabolism, including acetyl coenzyme A, pantothenic acid, L-tyrosine, threonyltyrosine, leucylproline, and Ser-Met.

3.4. Functional enrichment analysis of differentially metabolites

The biological pathways were annotated and enriched, associated with the identified differential metabolites, using the KEGG pathway analysis (Fig. S2A). The differential metabolites were mainly involved in pathways closely related to muscle development, including aminoacyl-tRNA biosynthesis, biosynthesis of amino acids, protein digestion and absorption, and purine metabolism.

To characterize coordinated shifts among metabolites, we calculated pairwise Pearson correlation coefficients for all detected metabolites (Fig. S2B). The resulting correlation matrix highlights several metabolites closely associated with muscle metabolism. Tryptophan and acetyl-coenzyme A, both essential for energy production and protein synthesis, form a positively correlated cluster. In contrast, tetradecanoic acid and aspartate, which participate in lipid utilization and amino acid turnover, exhibit strong negative correlations with this group.

3.5. Metabolite profile analysis of HSSPME-GC-MS

To characterize volatile compounds between MG and HB geese, HSSPME-GC-MS was performed on goose breast meat. The qualitative and quantitative data of volatile metabolites are summarized in Table S4. The TIC of mixed samples and the overlaid TICs of QC samples are shown in Fig. S3, indicating good analytical stability and reproducibility of the GC-MS system. A total of 821 volatile compounds were detected, annotated, and classified into 11 chemical classes (Fig. 5B). Among them, aldehydes, ketones, and esters were the most abundant group (30.21%), followed by hydrocarbons (23.75%) and alcohols and amines (16.69%). In addition, 74 heterocyclic compounds and 54 terpenoids were detected. The remaining classes included 46 benzenes and substituted derivatives, 29 organic acids and their derivatives, 16 ethers, 11 nitrogen-containing compounds, 7 halogenated hydrocarbons, and 4 other compounds. Relative odour activity value (rOAV) was used to estimate the potential contributions of volatile compounds to aroma based on literature-reported odour thresholds. The rOAV scatter plot is presented in Fig. 5A.

Fig. 5.

Fig. 5

Open in a new tab

Volatile metabolite profiling of breast muscle from MG and HB geese by HSSPME-GC-MS. (A) Scatter plot of relative odour activity values (rOAVs) showing the estimated potential contribution of volatile compounds to aroma. (B) Donut chart illustrating the chemical class distribution of annotated volatile compounds. (C) OPLS-DA score plot demonstrating the discrimination of volatile metabolite profiles between MG and HB geese (n = 6). (D) OPLS-DA permutation test confirming model validity. (E) Radar plot presenting the relative abundance patterns of significantly differential volatile metabolites between the two breeds. (F) Correlation network constructed based on differential metabolites and VOCs using Spearman's correlation analysis.

We performed pairwise comparisons using an OPLS-DA model (Fig. 5C and D), which showed good model performance (R2X = 0.749, R2Y = 0.914, p = 0.13) and predictive ability (Q2 = 0.615, p = 0.015). Metabolites with a fold change ≥2 (upregulated) or ≤ 0.5 (downregulated) and a VIP value ≥1 were defined as significantly differential metabolites. FDR correction was also performed, and the adjusted p-values are provided in Table S5. Notably, among these differential metabolites, only three were downregulated, whereas twenty-one were upregulated in the MG group compared with the HB group.

Among the differential volatile compounds, 6-hepten-3-one, 4-methyl-cyclohexanemethanol, and 3,5-dimethyl-4-octanone were significantly downregulated in the MG group compared with the HB group, all of which are mainly lipid oxidation–related ketones and alcohols. In contrast, twenty-one metabolites were significantly upregulated in the MG group, predominantly including representative hydrocarbons, alcohols, ketones, lactones, phenols, and organic acid esters, such as trimethyl-octane, nonen-2-ol, methyl-furanone, methyl-hexenone, methoxy-phenol, and dimethyl succinate. These changes indicate pronounced breed-related differences in lipid metabolism and volatile compound composition in raw goose meat. As shown in Fig. 5E, compounds associated with estery and fruity odour descriptors showed relatively higher predicted contributions than those associated with nutty, green, floral, coconut, and gum notes. Quantitative descriptive sensory analysis showed no significant differences between MG and HB geese in bloody, pungent, fishy, rotten-egg, or ammonia-like odour intensity (P > 0.05; Table S6).

We further constructed a correlation network to integrate metabolomics and volatilomics data (Fig. 5F). The correlation network showed that N,N-dibutyl-formamide acted as a central hub, exhibiting strong correlations with multiple metabolites, including L-tyrosine, L-tryptophan, isoleucylproline, leucylproline, and ribonalactone. In addition, 3,5-dimethyl-4-octanone was associated with metabolites such as acetyl coenzyme A.

Overall, HSSPME-GC-MS analysis of MG and HB goose breast meat detected and annotated 821 volatile compounds across 11 chemical classes, with aldehydes, ketones, and esters being the most abundant. Using rOAV, the potential aroma contributions of volatile compounds were estimated, and differential metabolite analysis showed that the MG group had 21 upregulated and 3 downregulated compounds compared with the HB group. Downregulated metabolites were mainly lipid oxidation-related ketones and alcohols, while upregulated ones included hydrocarbons, alcohols, ketones, lactones, phenols, and esters.

4. Discussion

Goose meat quality is strongly shaped by genetic background, which governs muscle development patterns, nutrient deposition, and metabolic regulation. China's goose industry is dominated by indigenous breeds, while a limited number of Western improved breeds have been introduced for commercial production. Magang geese and Hungarian geese represent typical examples of these two genetic backgrounds and exhibit clear differences in growth performance, muscle traits, and meat quality.

In the present study, physicochemical analyses showed that HB geese had higher IMF content and lower pressing loss in both breast and leg muscles, whereas MG geese exhibited higher protein and moisture content, particularly in the leg muscle. These observations are consistent with previous studies in poultry, where fast-growing or improved genotypes accumulate more intramuscular fat while retaining water more efficiently (Huo et al., 2021), whereas indigenous or slow-growing lines maintain leaner muscle characteristics (Sirri et al., 2011). Muscle fiber morphology provides a potential structural basis for these differences. The leg muscles of both breeds displayed larger fiber diameters and higher shear force compared to breast muscles, and HB leg fibers were significantly thicker than those of MG. Although larger fiber diameter has been associated with reduced water loss in some cases, the relationship between muscle fiber characteristics and WHC is complex and influenced by multiple factors, including muscle structural organization and extracellular space. Therefore, the superior WHC observed in HB meat may be related not only to fiber size but also to differences in overall muscle microstructure, consistent with previous studies in poultry (Mo et al., 2023; Roy & Bruce, 2024).

Metabolomic analysis further revealed pronounced breed-dependent biochemical differences. HB breast muscle was enriched in lipid-related metabolites, including phospholipids such as LysoPC (P-18:1/0:0), 1-stearoyl-2-arachidonoyl-PS, and dihomo-γ-linolenic acid ethyl ester, consistent with its higher IMF deposition (Kodde et al., 2021). This is in line with reports in chicken and duck meat, where breeds or lines with higher intramuscular fat exhibit elevated phospholipid and fatty acid derivatives, which may influence flavor-related compound profiles (Chen, Cao, Chang, et al., 2023; Yang et al., 2025). In contrast, MG breast muscle contained higher levels of amino acids and nucleotide metabolites, including L-arginine, L-proline, balenine, inosine, uric acid, and S-adenosylhomocysteine. These compounds are critical in protein turnover, purine metabolism, and energy homeostasis (Wu et al., 2011), and have been reported to be associated with flavor-related chemical characteristics in meat (Hossain et al., 2024). The elevated inosine and uric acid in MG is particularly noteworthy because purine metabolites have been associated with flavor-related characteristics in poultry and waterfowl (Huang et al., 2022). Pathway enrichment analysis of differential metabolites highlighted amino acid metabolism, purine metabolism, and carbohydrate metabolism as the most affected pathways between breeds. These metabolic shifts are consistent with genotype-dependent differences observed in other poultry species, where energy metabolism and amino acid turnover influence both nutritional composition and flavor-related metabolite profiles (Mir et al., 2017; Wang et al., 2024). In MG, the coordinated upregulation of amino acids such as L-arginine and L-proline, alongside purine metabolites including inosine, uric acid, and S-adenosylhomocysteine, suggests differences in amino acid and nucleotide metabolism, which may contribute to breed-related differences in flavor-related chemical profiles. In contrast, HB meat exhibited a metabolic profile characterized by higher levels of lipid-related metabolites, particularly phospholipids, which may be related to increased IMF content and differences in meat quality traits.

Volatile metabolite profiling of raw breast muscle using HSSPME-GC-MS provided complementary information on breed-related differences in volatile profiles (Lee et al., 2015; W. Zhang et al., 2024). A total of 821 volatile compounds were detected and annotated, predominantly aldehydes, ketones, esters, and alcohols. MG breast muscle exhibited 21 upregulated volatiles, including esters (dimethyl succinate, methyl-furanone), alcohols (nonen-2-ol), lactones, and phenolic compounds, while three lipid oxidation-derived ketones and alcohols (6-hepten-3-one, 3,5-dimethyl-4-octanone, 4-methyl-cyclohexanemethanol) were downregulated. The increased abundance of esters and alcohols may indicate a greater potential contribution of compounds associated with fruity and sweet odour descriptors, whereas the lower abundance of some lipid oxidation-derived compounds may indicate differences in off-flavor-related volatile profiles in raw meat samples (Cadwallader & Singh, 2009; Shahidi & Hossain, 2022). These findings are consistent with reports in chicken (Cartoni Mancinelli et al., 2021) and livestock (Man et al., 2023), where breed- or genotype-dependent differences in lipid composition and oxidative susceptibility significantly affect VOC formation. Notably, the VOC patterns were consistent with non-volatile lipid metabolite profiles, suggesting that breed-specific differences in lipid metabolism may contribute to variations in both non-volatile and volatile compound profiles in goose meat. Integrating the metabolomic and volatilomic data, our results suggest that HB geese exhibit a metabolic profile characterized by increased lipid-related metabolites and energy storage, which may be associated with higher IMF content and water-holding capacity, whereas MG geese show higher levels of amino acids and nucleotide-related metabolites that may be relevant to flavor-related chemical characteristics. These metabolic differences are consistent with the observed variations in meat quality traits, where HB meat showed higher IMF content and lower pressing loss, while MG meat exhibited higher protein and moisture content. Such integrated analysis underscores the importance of considering both non-volatile and volatile metabolites when assessing breed-dependent meat quality.

Overall, this study demonstrates that breed-specific differences in muscle microstructure, amino acid metabolism, purine metabolism, and lipid metabolism are associated with variations in goose meat quality and flavor-related compounds. Our findings provide a basis for breed selection, nutrition management, and quality-oriented production in the goose industry. However, no significant breed differences were detected in the five aroma descriptors evaluated in raw breast muscle. Because cooking may substantially alter volatile profiles and sensory perception, the potential flavor contributions inferred from volatile compounds and rOAVs require further validation using cooked goose meat and sensory evaluation.

5. Conclusions

Taken together, this study demonstrates that breed strongly influences goose meat quality through coordinated regulation of muscle structure, nutrient deposition, and metabolic pathways. Hungarian geese exhibited higher intramuscular fat and lower pressing loss, reflecting enhanced lipid accumulation and water-holding capacity, whereas Magang geese showed elevated protein and moisture contents. Integrated metabolomic and volatilomic analyses revealed that HB meat was enriched in lipid-related metabolites, particularly phospholipids, while MG meat contained higher levels of amino acids, nucleotides, and organic acids, which may contribute to breed-related differences in flavor-related chemical profiles. Volatile profiling of raw breast muscle further revealed breed-specific differences in compounds predicted to contribute to flavor, with MG displaying higher abundances of esters and alcohols and reduced lipid oxidation-derived compounds. Our findings provide insights into the metabolic characteristics of different goose breeds and may offer useful references for breed selection, nutritional management, and quality-oriented production in the goose industry.

CRediT authorship contribution statement

Kaiqi Weng: Writing – original draft, Investigation, Formal analysis, Data curation. Yi Liu: Validation, Supervision, Project administration, Data curation, Conceptualization. Huiying Wang: Supervision, Software. Yu Tan: Resources. Daqian He: Writing – review & editing, Supervision, Funding acquisition.

Ethics declaration

Animal subject

This study was conducted in accordance with the ARRIVE (Animal Research: Reporting of In Vivo Experiments) guidelines.

This study was approved by the the Institutional Animal Care and Use Committee of the Shanghai Academy of Agricultural Sciences.

(Approval No. SAASXM0524061)

Funding

This work was supported by the Shanghai “Science and Technology Innovation Action Plan” Morning Star Cultivation (Sailing Program 24YF2737800), the Agriculture Research System of China of MOF and MARA (CARS-41-7) and Shanghai Academy of Agricultural Sciences “Shang Nong Ying Cai” Cultivation Project: “Zhu Pao” Program and “Jie Bang Gua Shuai” Program (ZP26171 and BS25017).

Declaration of competing interest

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

Footnotes

Appendix A

Supplementary data to this article can be found online at https://doi.org/10.1016/j.fochx.2026.104417.

Appendix A. Supplementary data

Supplementary Figure S1.

Supplementary Figure S1

Open in a new tab

Quality control (QC) assessment of LC-MS analytical stability and reproducibility. (A and B) Total ion chromatogram (TIC) alignment of QC samples in positive and negative ion modes, respectively. (C and D) Distribution of relative standard deviation for detected features in positive and negative ion modes, respectively.

Supplementary Figure S2.

Supplementary Figure S2

Open in a new tab

Functional enrichment analysis of differentially metabolites. (A) KEGG pathway analysis of the identified differential metabolites. (B) Pearson correlation coefficients for all detected metabolites.

Supplementary Figure S3.

Supplementary Figure S3

Open in a new tab

GC-MS total ion chromatogram (TIC) profiles for quality assessment. (A) TIC of mixed samples. (B) Overlaid TICs of QC samples.

Supplementary Table S1

Ingredient and nutrient levels of the commercial diets in geese (1～70 d).

mmc4.docx (17.4KB, docx)
Supplementary Table S2

LC-MS/MS identification and quality control information of metabolites detected in raw breast meat from Magang and Hungarian geese.

mmc5.xlsx (13.5MB, xlsx)
Supplementary Table S3

Differential metabolites identified by LC-MS in raw breast meat from Magang and Hungarian geese.

mmc6.xls (539KB, xls)
Supplementary Table S4

Table S4. Qualitative and quantitative data of volatile compounds detected in raw breast meat from Magang and Hungarian geese.

mmc7.xlsx (320KB, xlsx)
Supplementary Table S5

Table S5. Differential volatile compounds identified by HSSPME-GC-MS in raw breast meat from Magang and Hungarian geese.

mmc8.xlsx (245KB, xlsx)
Supplementary Table S6

Table S6. Comparison of descriptive sensory intensity of raw breast meat between Magang and Hungarian geese.

mmc9.docx (17.5KB, docx)

Data availability

Data will be made available on request.

References

  1. Bueno M., Resconi V.C., Campo M.M., Ferreira V., Escudero A. Development of a robust HS-SPME-GC-MS method for the analysis of solid food samples. Analysis of volatile compounds in fresh raw beef of differing lipid oxidation degrees. Food Chemistry. 2019;281:49–56. doi: 10.1016/j.foodchem.2018.12.082. [DOI] [PubMed] [Google Scholar]
  2. Cadwallader K.R., Singh T.K. In: Advanced dairy chemistry: Volume 3: Lactose, water, salts and minor constituents. McSweeney P., Fox P.F., editors. Springer New York; New York, NY: 2009. Flavours and off-Flavours in Milk and dairy products; pp. 631–690. [Google Scholar]
  3. Cartoni Mancinelli A., Silletti E., Mattioli S., Dal Bosco A., Sebastiani B., Menchetti L.…Castellini C. Fatty acid profile, oxidative status, and content of volatile organic compounds in raw and cooked meat of different chicken strains. Poultry Science. 2021;100(2):1273–1282. doi: 10.1016/j.psj.2020.10.030. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Chen G., Qi L., Zhang S., Peng H., Lin Z., Zhang X.…Luo W. Metabolomic, lipidomic, and proteomic profiles provide insights on meat quality differences between Shitou and Wuzong geese. Food Chemistry. 2024;438 doi: 10.1016/j.foodchem.2023.137967. [DOI] [PubMed] [Google Scholar]
  5. Chen X., Cao J., Chang C., Geng A., Wang H., Chu Q.…Zhang J. Effects of age on compounds, metabolites and meat quality in Beijing-you chicken breast meat. Animals : an Open Access Journal from MDPI. 2023;13 doi: 10.3390/ani13213419. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Chen X., Cao J., Geng A., Zhang X., Wang H., Chu Q.…Zhang J. Integration of GC-MS and LC-MS for metabolite characteristics of thigh meat between fast- and slow-growing broilers at marketable age. Food Chemistry. 2023;403 doi: 10.1016/j.foodchem.2022.134362. [DOI] [PubMed] [Google Scholar]
  7. Chong J., Xia J. MetaboAnalystR: An R package for flexible and reproducible analysis of metabolomics data. Bioinformatics. 2018;34(24):4313–4314. doi: 10.1093/bioinformatics/bty528. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. FAO-STAT 2023. https://www.fao.org/faostat/en/#data/QCL
  9. Fiehn O. Metabolomics by gas chromatography-mass spectrometry: Combined targeted and untargeted profiling. Current Protocols in Molecular Biology. 2016;114 doi: 10.1002/0471142727.mb3004s114. 30.34.31–30.34.32. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Guo Y., Cao Z., Weng K., Zhang Y., Zhang Y., Chen G., Xu Q. Effect of chilled storage period on the volatile organic compounds and bacterial community in goose meat. Food Chem X. 2024;23 doi: 10.1016/j.fochx.2024.101685. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Hossain M.J., Alam A.N., Lee E.Y., Hwang Y.H., Joo S.T. Umami characteristics and taste improvement mechanism of meat. Food Science of Animal Resources. 2024;44(3):515–532. doi: 10.5851/kosfa.2024.e29. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Huang Z., Zhang J., Gu Y., Cai Z., Feng X., Yang C., Xin G. Research progress on inosine monophosphate deposition mechanism in chicken muscle. Critical Reviews in Food Science and Nutrition. 2022;62(4):1062–1078. doi: 10.1080/10408398.2020.1833832. [DOI] [PubMed] [Google Scholar]
  13. Huo W., Weng K., Gu T., Zhang Y., Zhang Y., Chen G., Xu Q. Effect of muscle fiber characteristics on meat quality in fast- and slow-growing ducks. Poultry Science. 2021;100(8) doi: 10.1016/j.psj.2021.101264. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Ismail I., Joo S.T. Poultry meat quality in relation to muscle growth and muscle fiber characteristics. Korean Journal for Food Science of Animal Resources. 2017;37(6):873–883. doi: 10.5851/kosfa.2017.37.6.87. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Joo S.T., Kim G.D., Hwang Y.H., Ryu Y.C. Control of fresh meat quality through manipulation of muscle fiber characteristics. Meat Science. 2013;95(4):828–836. doi: 10.1016/j.meatsci.2013.04.044. [DOI] [PubMed] [Google Scholar]
  16. Kanehisa M., Goto S. KEGG: Kyoto encyclopedia of genes and genomes. Nucleic Acids Research. 2000;28(1):27–30. doi: 10.1093/nar/28.1.27. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Katiyo W., de Kock H.L., Coorey R., Buys E.M. Sensory implications of chicken meat spoilage in relation to microbial and physicochemical characteristics during refrigerated storage. LWT. 2020;128 doi: 10.1016/j.lwt.2020.109468. [DOI] [Google Scholar]
  18. Kodde A., Mischke M., Rakhshandehroo M., Voggel J., Fink G., Nüsken E.…Nüsken K.-D. The effect of dietary lipid quality in early life on serum LysoPC(18,2) levels and their association with adult blood glucose levels in intrauterine growth restricted rats. Nutrition & Metabolism. 2021;18(1):101. doi: 10.1186/s12986-021-00614-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Krunt O., Zita L. Goose meat production and consumption: Current knowledge and market challenges. World's Poultry Science Journal. 2025:1–13. doi: 10.1080/00439339.2025.2564803. [DOI] [Google Scholar]
  20. Lee C., Lee Y., Lee J.-G., Buglass A.J. Improving the extraction of headspace volatile compounds: Development of a headspace multi-temperature solid-phase micro-extraction-single shot-gas chromatography/mass spectrometry (mT-HSSPME-ss-GC/MS) method and application to characterization of ground coffee aroma. Analytical Methods. 2015;7(8):3521–3536. doi: 10.1039/C4AY03034F. [DOI] [Google Scholar]
  21. Longo V., Forleo A., Provenzano S.P., Coppola L., Zara V., Ferramosca A.…Capone S. HS-SPME-GC-MS metabolomics approach for sperm quality evaluation by semen volatile organic compounds (VOCs) analysis. Biomedical Physics & Engineering Express. 2019;5(1) doi: 10.1088/2057-1976/aaeb07. [DOI] [Google Scholar]
  22. Man L., Ren W., Qin H., Sun M., Yuan S., Zhu M.…Li M. Characterization of the relationship between lipids and volatile compounds in donkey, bovine, and sheep meat by UHPLC–ESI–MS and SPME–GC–MS. LWT. 2023;175 doi: 10.1016/j.lwt.2023.114426. [DOI] [Google Scholar]
  23. Mir N.A., Rafiq A., Kumar F., Singh V., Shukla V. Determinants of broiler chicken meat quality and factors affecting them: A review. Journal of Food Science and Technology. 2017;54(10):2997–3009. doi: 10.1007/s13197-017-2789-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Mo M., Zhang Z., Wang X., Shen W., Zhang L., Lin S. Molecular mechanisms underlying the impact of muscle fiber types on meat quality in livestock and poultry. Frontiers in Veterinary Science. 2023;10 doi: 10.3389/fvets.2023.1284551. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Muroya S., Ueda S., Komatsu T., Miyakawa T., Ertbjerg P. MEATabolomics: Muscle and meat metabolomics in domestic animals. Metabolites. 2020;10(5) doi: 10.3390/metabo10050188. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Roy B.C., Bruce H.L. Contribution of intramuscular connective tissue and its structural components on meat tenderness-revisited: A review. Critical Reviews in Food Science and Nutrition. 2024;64(25):9280–9310. doi: 10.1080/10408398.2023.2211671. [DOI] [PubMed] [Google Scholar]
  27. Shahidi F., Hossain A. Role of lipids in food flavor generation. Molecules. 2022;27 doi: 10.3390/molecules27155014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Sirri F., Castellini C., Bianchi M., Petracci M., Meluzzi A., Franchini A. Effect of fast-, medium- and slow-growing strains on meat quality of chickens reared under the organic farming method. Animal. 2011;5(2):312–319. doi: 10.1017/S175173111000176X. [DOI] [PubMed] [Google Scholar]
  29. Tan J., Mu B., Li H., Li C., Gao T., Meng X.…Li G. A comprehensive study of meat quality and flavor characteristics of different sexes of Yanbian yellow cattle using GC-IMS and LC-MS/MS technologies. Foods. 2025;14(18) doi: 10.3390/foods14183175. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Wang Y., Liu L., Liu X., Wang Y., Yang W., Zhao W.…Wen J. Identification of characteristic aroma compounds in chicken meat and their metabolic mechanisms using gas chromatography–olfactometry, odor activity values, and metabolomics. Food Research International. 2024;175 doi: 10.1016/j.foodres.2023.113782. [DOI] [PubMed] [Google Scholar]
  31. Weng K., Huo W., Song L., Cao Z., Zhang Y., Zhang Y.…Xu Q. Effect of marketable age on nutritive profile of goose meat based on widely targeted metabolomics. LWT. 2022;170 doi: 10.1016/j.lwt.2022.114071. [DOI] [Google Scholar]
  32. Wu G., Bazer F.W., Burghardt R.C., Johnson G.A., Kim S.W., Knabe D.A.…Spencer T.E. Proline and hydroxyproline metabolism: Implications for animal and human nutrition. Amino Acids. 2011;40(4):1053–1063. doi: 10.1007/s00726-010-0715-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Yang M., Liu X., Li R., Liang Z., Wang Q., Kong Y.…Huang H. Differences in flavour compounds and key metabolic markers in high-quality broiler rooster breast muscle based on broad-target metabolomics and volatile metabolomics. Foods. 2025;14 doi: 10.3390/foods14234089. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Yuan H., Cao G., Hou X., Huang M., Du P., Tan T.…Luo J. Development of a widely targeted volatilomics method for profiling volatilomes in plants. Molecular Plant. 2022;15(1):189–202. doi: 10.1016/j.molp.2021.09.003. [DOI] [PubMed] [Google Scholar]
  35. Zhang W., Yu X., Xin L., Xu S., Meng X. Effect of rapid fermentation on the quality of northeastern sauerkraut analyzed based on HSSPME-GC-MS and LC-MS. LWT. 2024;198 doi: 10.1016/j.lwt.2024.116005. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Table S1

Ingredient and nutrient levels of the commercial diets in geese (1～70 d).

mmc4.docx (17.4KB, docx)
Supplementary Table S2

LC-MS/MS identification and quality control information of metabolites detected in raw breast meat from Magang and Hungarian geese.

mmc5.xlsx (13.5MB, xlsx)
Supplementary Table S3

Differential metabolites identified by LC-MS in raw breast meat from Magang and Hungarian geese.

mmc6.xls (539KB, xls)
Supplementary Table S4

Table S4. Qualitative and quantitative data of volatile compounds detected in raw breast meat from Magang and Hungarian geese.

mmc7.xlsx (320KB, xlsx)
Supplementary Table S5

Table S5. Differential volatile compounds identified by HSSPME-GC-MS in raw breast meat from Magang and Hungarian geese.

mmc8.xlsx (245KB, xlsx)
Supplementary Table S6

Table S6. Comparison of descriptive sensory intensity of raw breast meat between Magang and Hungarian geese.

mmc9.docx (17.5KB, docx)

Data Availability Statement

Data will be made available on request.

Articles from Food Chemistry: X are provided here courtesy of Elsevier

RESOURCES