Goats (Capra hircus), which were domesticated around 10,000 years ago in the Fertile Crescent (Zeder and Hesse, Reference Zeder and Hesse2000), now number more than one billion worldwide. 95% of them live in Asia, Africa and Latin America (Niyazbekova et al., Reference Niyazbekova, Yao, Liu, Bold, Tong, Chang, Wen, Li, Wang, Chen and Ma2020). In China, Guanzhong dairy goats are one of the most productive indigenous breeds and play a key role in the regional development of the dairy industry. Despite their economic importance, the detailed profiling of milk in early lactation, especially the transition from colostrum (CM) to transition milk (TM), is still limited in this breed.
Colostrum, secreted within the first 24 h postpartum, is uniquely enriched with immunoglobulins (IgG, IgA and IgM), growth factors and oligosaccharides, which confer passive immunity and support gut maturation (Sánchez-Macías et al., Reference Sánchez-Macías, Moreno-Indias, Castro, Morales-Delanuez and Argüello2014; Godden and Lombard, Reference Godden, Lombard and Woolums2019). In contrast, TM, produced within the first week post-partum, marks a gradual shift towards increased energy density mainly via fat and lactose and decreased immune components, aligning with the neonate's evolving physiological requirements (Fischer-Tlustos et al., Reference Fischer-Tlustos, Hertogs, Van Niekerk, Nagorske, Haines and Steele2020). In addition to their nutritional and immunological functions, CM and TM also serve as a medium for the transfer of microorganisms. Bacterial communities in CM, TM and milk can arise from external contamination of the teat or via endogenous pathways, including translocation from the gastrointestinal tract (Doyle et al., Reference Doyle, Gleeson, O'toole and Cotter2016; Porcellato et al., Reference Porcellato, Meisal, Bombelli and Narvhus2020). These bacteria can influence neonatal gut colonisation and reflect maternal health status (D'amico, Reference D'amico and Donnelly2010; Fecteau et al., Reference Fecteau, Baillargeon, Higgins, Paré and Fortin2002). At the same time, long-chain saturated fatty acids (FA; ≥C12), such as palmitic acid (C16:0) and stearic acid (C18:0), serve not only as important energy sources but also as modulators of host metabolic processes, including post-translational modifications (Legrand, Reference Legrand and Rioux2015; Ruiz-Núñez et al., Reference Ruiz-Núñez, Dijck-Brouwer and Muskiet2016). Of particular interest are odd- and branched-chain FA (OBCFAs), which are predominantly derived from microbial metabolism in the rumen and have been associated with the maintenance of gut health and metabolic stability in ruminants (Vlaeminck et al., Reference Vlaeminck, Fievez, Cabrita, Fonseca and Dewhurst2006; Gómez-Cortés and Juárez, Reference Gómez-Cortés, Juárez and De La Fuente2018).
This study addresses critical knowledge gaps by characterising the compositional and functional shifts in milk between CM and TM in Guanzhong dairy goats. Specifically, we aimed to investigate changes in immunoglobulin concentrations, long-chain saturated FA profiles, and microbial community structure, with a focus on identifying relationships among these components. Through an integrated analysis of immune, lipid and microbial dynamics during early lactation, this study contributes to a deeper understanding of goat milk biology and supports the development of evidence-based strategies for colostrum and transition milk management to improve neonatal health.
Material and methods
All animal handling procedures were conducted in accordance with the experimental protocols approved by the Animal Care Committee (ACC) of the Inner Mongolia Academy of Agricultural & Animal Husbandry Sciences (Protocol number: 2021–031). All ACC guidelines and methods were followed throughout the experiment.
Animals, diets and sample collection
Milk samples were collected from eight one-year-old primiparous Guanzhong dairy goats on a local farm (Huayile Co., Ltd, Hohhot, Inner Mongolia, China, 40.88 N 111.70 E) from March 8 to 20, 2021. All goats were kept indoors on wheat straw bedding during the transition period. Dietary ingredients and nutritional compositions are shown in Table 1. The diets were formulated according to a domestic feeding standard to ensure an average milk production of over 1.37 kg/day (500 kg/year), and feed intake averaged 2.0 kg/d. None of the goats required antibiotics or other treatment which may affect the milk microbial composition, and all udder halves were healthy during the sampling period. Milk samples were collected from each animal immediately (colostrum, CM; n = 8) and then on day 7 (transition milk, TM; n = 8) after kidding. Before sampling, the goats were restrained and the teats were sterilised with a 0.5% iodine disinfectant. Fifteen mL of iodine was applied to each teat. The sampler wore nitrile gloves, which were changed between animals. Thirty mL of the hand-collected milk samples from each goat were aseptically poured into three sterile 10 mL tubes (the first 3 drops of milk were discarded to avoid contamination with disinfectant) and immediately placed in liquid nitrogen. The samples were then stored in a freezer at −80°C until further analysis. During the sample collection period, all kids were kept with their dams and suckled colostrum and transition milk freely.
Ingredients and chemical composition of experimental diets
1 The premix provided the following per kg of diets: VA 154,440 IU, VD3 94,000 IU, VE 338,200, I 120 mg, Cu 280 mg, Fe 2240 mg, Mn 1740 mg, Zn 1370 mg, Se 60 mg and Co 16.8 mg.
Measurements of immunoglobulins
Immunoglobulin concentrations in goat CM and TM samples (10 mL) were measured using ELISA kits specific for goat IgG (Product No.: 25,602), IgM (Product No.: 25,607) and IgA (Product No.: 25,601), following the manufacturer's instructions (mlbio, Shanghai Enzyme-linked Biotechnology Co., Ltd., Shanghai, China).
Measurement of long-chain saturated FA
Long-chain saturated FA was quantified using a Shimadzu GC-2010 gas chromatograph coupled to an Equity™-1 capillary column (15 m × 0.1 μm). The injection port and detector were maintained at 280°C. The temperature program started at 175°C, followed by an increase to 275°C at a rate of 15°C/min, with a 1-minute hold time to facilitate separation of the fatty acid methyl esters. The carrier gas, nitrogen, was set to a flow rate of 45 cm/s, while hydrogen and air were supplied at flow rates of 40 mL/min and 400 mL/min, respectively.
Microbial DNA extraction and amplicon sequencing of milk bacteria
Total milk bacteria DNA was extracted using the PowerSoil DNA Isolation Kit (MoBio Laboratories, Carlsbad, CA) according to the manufacturer's guidelines. Quantity and quality of DNA was determined using a Nanodrop 1000 spectrophotometer (Thermo Scientific, Shanghai, China). To analyse the bacterial profiles, the hypervariable region V3-V4 of the 16S rRNA genes was amplified with the following primers: 338 F (5′-ACTCCTACGGGAGGCAGCAG-3′) and 806 R (5′-GGACTACHVGGGTWTCTAAT-3′). The PCR products were purified using an Agencourt AMPure XP kit (Beckman Coulter, Shanghai, China). The quality and quantity of the purified amplicon was determined with the NanoDrop 1000 (NanoDrop Technologies, Wilmington, DE) and the PicofluorTM Handheld Fluorometer (ThermoFisher Scientific, Beijing, China) using Quant-iTTM PicoGreenTM dsDNA Reagents (ThermoFisher Scientific, Beijing, China). Amplicon sequencing with a pair-end was performed at Beijing Allwegene Tech Ltd (Beijing, China) using the Illumina MiSeq platform (2 × 300 bp; Illumina, San Diego, CA).
The sequencing data were analysed using the QIIME2 program (Bolyen et al., Reference Bolyen, Rideout, Dillon, Bokulich, Abnet, Al-Ghalith, Alexander, Alm, Arumugam, Asnicar, Bai, Bisanz, Bittinger, Brejnrod, Brislawn, Brown, Callahan, Caraballo-Rodríguez, Chase, Cope, Da Silva, Diener, Dorrestein, Douglas, Durall, Duvallet, Edwardson, Ernst, Estaki, Fouquier, Gauglitz, Gibbons, Gibson, Gonzalez, Gorlick, Guo, Hillmann, Holmes, Holste, Huttenhower, Huttley, Janssen, Jarmusch, Jiang, Kaehler, Kang, Keefe, Keim, Kelley, Knights, Koester, Kosciolek, Kreps, Langille, Lee, Ley, Liu, Loftfield, Lozupone, Maher, Marotz, Martin, McDonald, McIver, Melnik, Metcalf, Morgan, Morton, Naimey, Navas-Molina, Nothias, Orchanian, Pearson, Peoples, Petras, Preuss, Pruesse, Rasmussen, Rivers, Robeson, Rosenthal, Segata, Shaffer, Shiffer, Sinha, Song, Spear, Swafford, Thompson, Torres, Trinh, Tripathi, Turnbaugh, Ul-Hasan, van der Hooft, Vargas, Vázquez-Baeza, Vogtmann, von Hippel, Walters, Wan, Wang, Warren, Weber, Williamson, Willis, Xu, Zaneveld, Zhang, Zhu, Knight and Caporaso2019). Specifically, paired sequences were demultiplexed using the ‘q2-demux’ plugin before being quality checked using the ‘q2-dada2’ plugin (Callahan et al., Reference Callahan, Mcmurdie, Rosen, Han, Johnson and Holmes2016). Taxonomy was assigned to amplicon sequence variants (ASVs) using a pre-trained QIIME-compatible SILVA database (v138) with 99% identity and taxonomy was assigned to representative sequences using the ‘q2-feature-classifier’ plugin (Bokulich et al., Reference Bokulich, Kaehler, Rideout, Dillon, Bolyen, Knight, Huttley and Gregory Caporaso2018). The potentially contaminating sequences were removed using the ‘decontam’ package (version 1.12; Davis et al., Reference Davis, Proctor, Holmes, Relman and Callahan2018) according to the developer's instructions in R (R Core Team, 2020) using R Studio (Reference RacineRacine,).
To further investigate the function of the milk microbiota, the functionality (protein-coding genes) and pathways were predicted using the ‘q2-pictust2’ plugin (Douglas et al., Reference Douglas, Maffei, Zaneveld, Yurgel, Brown, Taylor, Huttenhower and Langille2020). The predicted Kyoto Encyclopedia of Genes and Genomes (KEGG) orthologs that did not meet the following criteria were removed: a minimum number of 10, a prevalence of 20% in the samples and 10% of the samples below the interquartile range.
Data calculation and statistical analyses
Principal coordinate analysis (PCoA) based on Bray–Curtis distance was performed using R. Specifically, the ‘vegdist’ function in ‘vegan’ package (version 2.7–1) was used to calculate the Bray–Curtis distance matrix, ‘cmdscale’ function in ‘stats’ package (version 4.4.0) was used to perform PCoA analysis, and ‘adonis2’ function in ‘vegan’ package (version 2.7–1) was used for Permutational multivariate analysis of variance (PERMANOVA). All of the above analyses were performed in R. The concentrations of IgG, IgM and IgA were analysed using estimation methods, with the results presented as mean difference estimation plots (Claridge-Chang and Assam, Reference Claridge-Chang and Assam2016). Effect sizes were calculated using Hedges’ g (Greenland et al., Reference Greenland, Senn, Rothman, Carlin, Poole, Goodman and Altman2016) and classified as trivial (g < 0.2), small (0.2 < g < 0.5), moderate (0.5 < g < 0.8) or large (g > 0.8). These effect sizes were presented as bootstrap 95% confidence intervals on aligned axes. Mean differences were estimated using bootstrap methods (resampled 5,000 times, bias-corrected and accelerated) and displayed with the bootstrap distribution of the mean. P-values were calculated using the Mann–Whitney U statistic for reporting purposes (Crichton, Reference Crichton2000). Differences in Shannon index and relative abundance of bacteria and relative abundance of KEGG orthologs and pathways between CM and TM samples were assessed using the non-parametric Wilcoxon rank sum test (Wilcoxon, Reference Wilcoxon and Johnson1992) using the ‘wilcox.test’ function of the ‘stats’ package (version 4.4.0) in R. Significant difference was declared at P ≤ 0.05 and trends at 0.05 < P ≤ 0.10.
The Spearman correlation between bacteria and FA was performed using the ‘rcorr’ function of Hmsic package (version 5.2–3) in R. Spearman's correlation coefficient, known as rho (ρ), ranges from −1.00 (a perfect negative correlation) to + 1.00 (a perfect positive correlation). A rho value greater than 0.50 or less than −0.50 and a P value ≤0.05 is considered a significant correlation. Data were expressed as mean ± standard deviation.
Sequence accession numbers
All DNA sequences identified in this study have been deposited in the NCBI sequence read archive under the accession numbers SRR16214161 to SRR16214176.
Results
Concentrations of immunoglobulins and long-chain saturated FA
The concentrations of IgG (102.8 ± 23.5 vs. 44.5 ± 10.7 mg/mL), IgM (3.4 ± 0.7 vs. 1.5 ± 0.3 mg/mL) and IgA (4.2 ± 0.8 vs. 1.7 ± 0.3 mg/mL) were significantly lower (P < 0.001) in TM compared to CM, with paired Hedges’ g values indicating a large effect size (Fig. 1).
Concentrations (mg/mL) of immunoglobulins G (IgG), A (IgA) and M (IgM) in goat colostrum (CM) and transition milk (TM) samples. The paired Hedges’ g values indicate the effect size between CM and TM, with negative values showing a decrease in concentration from CM to TM. Error bars represent standard deviation, and black circles indicate mean values. The line represents individual data points. Significant differences were observed for all immunoglobulins (P < 0.001).
The concentration of palmitic acid (C16:0) was the highest among all the samples, followed by stearic acid (C18:0), myristic acid (C14:0) and nonadecanoic acid (C19:0) (Table 2). Compared to that in CM, the concentration of total long-chain saturated FA (P = 0.015), total branch-chain FA (BCFA, P = 0.011), iso-branch-chain FA (isoBCFA, P = 0.022), odd-chain branched FA (OBCFA, P = 0.022) and even-branch-chain FA (even BCFA, P = 0.022) was significantly higher in TM. Among these, the concentration of FA iso-C15:0 (P = 0.017), C16:0 (P = 0.041), anteiso-C17:0 (P = 0.011), C17:0 (P = 0.005), iso-C18:0 (P = 0.025), C18:0 (P = 0.011) and C19:0 (P = 0.011) in CM were significantly lower.
Concentrations (g/100 g FA) of long-chain fatty acids in goat colostrum (CM) and transition milk (TM) samples
BCFA, branched-chain fatty acids; CM, colostrum; TM, transition milk.
1 iso-BCFA: iso-C14:0, iso-C15:0, iso-C16:0, iso-C17:0, iso-C18:0 and iso-C20:0.
2 anteiso-BCFA: anteiso-C13:0, anteiso-C15:0, anteiso-C17:0 and anteiso-C19:0.
3 Odd-chain BCFA: anteiso-C13:0, iso-C15:0, anteiso-C15:0, iso-C17:0, anteiso-C17:0 and anteiso-C19:0.
4 Even-chain BCFA: iso-C14:0, iso-C16:0, iso-C18:0, and iso-C20:0.
Diversity, taxonomic composition and predicted function of bacteria
A total of 1,105,195 high-quality 16S rRNA sequencing reads were generated for 16 milk samples. After quality control (QC) and removal of potential contaminants, the remaining 596,489 reads were assigned to 4,739 ASVs, with an average of 33,288 ± 2,659 (mean ± standard error) post-QC reads and 443 ± 49 ASVs per sample based on 99% nucleotide sequence similarity. The Chao 1 and Shannon index was similar between CM and TM samples (P = 0.28; Fig. 2a and 2b). The principal coordinate analysis (PCoA) based on Bray–Curtis distance showed that bacterial profiles did not differ between the two groups of samples (PERMANOVA P = 0.185; Fig. 2c).
Analysis of the bacterial diversity in goat colostrum (CM) and transition milk (TM) samples. Difference in the Chao1 index (a) and Shannon index (b). The principal coordinate analysis (PCoA) based on Bray–Curtis distance of microbial profiles at amplicon sequence variants (ASV) levels between colostrum (CM) and transition milk (TM) samples (c).
Fourteen bacterial phyla were identified in both CM and TM samples (relative abundance > 0.1% in at least 4 samples), predominant by Proteobacteria (59.0% ± 10.6%), Firmicutes (18.0% ± 10.5%) and Bacteroidetes (9.6% ± 2.6%) (Fig. 3a). A total of 62 bacterial genera were identified using the same cut-off mentioned above, predominant by Brevundimonas (12.6% ± 3.4%), Acinetobacter (11.1% ± 3.9%) and Sphingopyxis (9.8% ± 2.4) (Fig. 3a). Comparison analysis showed that the relative abundance of Salinicoccus, unclassified Lachnospiraceae, Atopostipes, Ruminococcus and Fusobacterium increased significantly (P < 0.05) in TM samples (Fig. 3b).
Taxonomic composition of bacteria in goat colostrum (CM) and transition milk (TM) at phylum and genus level (a). The relative abundance of 7 genera significantly different between goat colostrum (purple bar) and transition milk (green bar) samples (b).
The PICRUST2 was used to predict the protein-coding genes as well as functional pathways of milk bacteria. The main functions of bacteria in both CM and TM samples were associated with quorum sensing as well as amino and fatty acid metabolism (Fig. 4a). The relative abundance of several pathways was differentially abundant between CM and TM samples. For example, the relative abundance of ‘cold shock protein’ (K03704, P = 0.038) was higher in CM than in TM samples (Fig. 4b). In addition, the relative abundance of PWY-5918 (P < 0.05), PWY-5189 (P < 0.05), HEMESYN2-PWY (P = 0.05) and PWY-5188 (P = 0.05) was higher, while that of PWY-6545 (P < 0.05), PWY-7328 (P < 7328), PWY-6737 (P < 0.05), GLYCOAT-PWY (P < 0.05), P161-PWY (P < 0.05) and NONOXIPENT-PWY (P = 0.05) was lower in CM compared with TM samples (Fig. 4b).
Analysis of the predicted functions and pathways of microbiota in goat colostrum (CM) and transition milk (TM) samples using PICRUST2. Predicted functions of bacteria in goat CM and TM samples (a). Differential pathways in goat CM and TM samples (b).
Relationship between bacteria and long-chain saturated FA
We identified a significant correlation between several milk bacterial genera and the concentration of long-chain saturated FA. For example, the relative abundance of Fusobacterium was positively correlated with the concentration of iso-C15:0 (rho = 0.71, P = 0.004) anteiso-C17:0 (rho = 0.68, P = 0.007), C19:0 (rho = 0.67, P = 0.007) (Fig. 5). The relative abundance of Bosea was positively correlated with the concentration of iso-C17:0 (rho = 0.66, P = 0.010), anteiso-C17:0 (rho = 0.67, P = 0.009), C17:0 (rho = 0.65, P = 0.01) and anteiso-C19:0 (rho = 0.59, P = 0.025). The related abundance of unclassified Lachnospiraceae was positively correlated with the concentration of anteiso-C17:0 (rho = 0.64, P = 0.15) and C17:0 (rho = 0.59, P = 0.027). The related abundance of Streptococcus was positively correlated with the concentration of iso-C16:0 (rho = 0.59, P = 0.027), anteiso-C17:0 (rho = 0.74, P = 0.003), C17:0 (rho = 0.61, P = 0.02) and anteiso-C19:0 (rho = 0.62, P = 0.017). The relative abundance of Faecalibactrium was negatively correlated with the concentration of C13:0 (rho = −0.54, P = 0.447).
Spearman correlation between bacteria in goat colostrum and transition milk samples and long-chain saturated fatty acids.
Discussion
This study provides a detailed examination of the compositional dynamics between CM and TM in Guanzhong dairy goats, focusing on immunoglobulins, long-chain saturated FA, milk microbiota. Colostrum provides high levels of immunoglobulins critical for neonatal immunity (Godden and Lombard et al., Reference Godden, Lombard and Woolums2019), with a significant decline in IgG, IgM and IgA from CM to TM in this study. This shift reflects the primary role of colostrum in passive immunity, as documented in ruminants (Godden and Lombard et al., Reference Godden, Lombard and Woolums2019). The reduced immunoglobulin content in TM supports the evolving nutritional needs of the neonate, emphasizing the importance of timely colostrum feeding for neonatal health and survival (Godden and Lombard et al., Reference Godden, Lombard and Woolums2019).
In contrast to the immunoglobulins, the TM had significantly higher concentrations of total long-chain saturated FA, total BCFA and specific FA such as iso-C15:0, C16:0, anteiso-C17:0, C17:0 and C18:0. The FA hierarchy, dominated by C16:0 (palmitic acid), followed by C18:0 (stearic acid), C14:0 (myristic acid) and C19:0 (nonadecanoic acid), reflects a lipid profile tailored for energy supply and growth. This shift likely supports the increasing metabolic needs of the developing kid as it transitions from passive immunity to active metabolism. The BCFA that increased in TM may provide anti-inflammatory and metabolic benefits, as has been observed in other mammals (Ran-Ressler et al.,Reference Ran-Ressler, Bae, Lawrence, Wang and Brenna2014).
This study provides new insights into the microbial ecology of CM and TM in Guanzhong dairy goats and reveals a conserved core microbiota dominated by Proteobacteria, Firmicutes, Bacteroidota and Actinobacteria. These phyla together accounted for over 90% of the relative abundance, which is consistent with previous reports on bovine and goat milk (Wang et al., Reference Wang, Nan, Zhao, Wang, Wang, Jiang, Zhang, Xue, Hua, Li and Liu2020; Zhang et al., Reference Zhang, Wang, Lei, Wang, Jiang, Peng, Zhang and Shao2017; Bonsaglia et al., Reference Bonsaglia, Gomes, Canisso, Zhou, Lima, Rall, Oikonomou, Bicalho and Lima2017). Importantly, the detection of Sphingopyxis, a genus not previously detected in goat milk, adds a new dimension to current microbial inventories. Originally reclassified from Sphingomonas (Takeuchi and Hamana, Reference Takeuchi, Hamana and Hiraishi2001), Sphingopyxis is known for its ability to degrade xenobiotics (Sharma et al., Reference Sharma, Khurana, Singh and Negi2021), suggesting a possible role in the maintenance of mammary gland or milk hygiene through the degradation of environmental contaminants. While overall diversity metrics remained stable, shifts in taxonomic composition were observed across lactation stages. In particular, the TM samples showed increased abundances of Salinicoccus, Ruminococcus, Fusobacterium, Atopostipes and Lachnospiraceae. Several of these taxa, such as Ruminococcus and Lachnospiraceae, are known for their ability to produce short-chain FA (SCFA), particularly butyrate, which supports intestinal barrier integrity, immunomodulation and energy metabolism in neonates (Louis and Flint, Reference Louis and Flint2017; Martín et al., Reference Martín, Rios-Covian, Huillet, Auger, Khazaal, Bermúdez-Humarán, Sokol, Chatel and Langella2023). These changes in composition may reflect microbial adaptation to the evolving biochemical environment of the TM as it transitions from an immunological to a more nutritional function.
Functionally, the PICRUSt2 predictions showed different stage-specific microbial capacities. In CM, there was a greater enrichment of cold shock protein genes and heme biosynthetic pathways (e.g. PWY-5918, HEMESYN2-PWY), which could enhance bacterial resilience to post-kidding stress and contribute to the establishment of anaerobic conditions in the neonatal gut (Sturm-Richter et al., Reference Sturm-Richter, Golitsch, Sturm, Kipf, Dittrich, Beblawy, Kerzenmacher and Gescher2015; Hu et al., Reference Hu, Lu, Qin, Liu, Wang, Lu and Li2024). Such properties are potentially crucial during the early postnatal window, when milk serves not only as a vehicle for passive immunity but also for microbial colonisation of the gastrointestinal tract. In contrast, the TM microbiota exhibited increased representation of metabolic pathways related to starch and glycogen degradation (e.g. PWY-6737, GLYCOCAT-PWY), likely supporting the neonate's progressive adaptation to solid feed. These functional transitions reflect the ontogeny of the kid's digestive system and may reflect a coordinated interplay between host development and microbial succession. Overall, the predictive functional data are consistent with the hypothesis that the milk microbiota actively contributes to the changing physiological needs of the developing neonate.
The identification of genera such as Emticicia, which are significantly enriched in CM, emphasises the stage-dependent role of milk-associated bacteria. Emticicia oligotrophica, the type species of the genus, produces fucosidases capable of hydrolyzing fucosylated oligosaccharides, glycans found in high abundance in colostrum (Meyrand et al., Reference Meyrand, Dallas, Caillat, Bouvier, Martin and Barile2013; Liu et al., Reference Liu, Kulinich, Cai, Ma, Du, Lv and Liu2016). This enzymatic activity can improve the accessibility of nutrients and immune signalling in newborns. The enrichment of Salinicoccus, a halophilic genus previously observed in ruminant milk (Quigley et al., Reference Quigley, O'sullivan, Stanton, Beresford, Ross, Fitzgerald and Cotter2013; Zhang et al., Reference Zhang, Huo, Zhu and Mao2015), and Faecalibacterium, a known butyrate producer, in TM suggests a microbial shift towards taxa that can support energy metabolism and gut maturation. The identification of Faecalibacterium, a strictly anaerobic genus commonly originating from the gut (Young et al., Reference Young, Hine, Wallace, Callaghan and Bibiloni2015), raises the possibility of enteromammary transfer and warrants further investigation. Taken together, these data support the concept that the milk microbiota are not passive contaminants, but dynamically structured communities that reflect and potentially influence the biology of lactation and neonatal development.
In accordance with previous studies in dairy goats (Fan et al., Reference Fan, Zhang, Zhao, Ji, Du, Han, Wang and Yang2023; Ayala et al., Reference Ayala, Gómez-Cortés, Hernández, Madrid, MartÍnez-Miró and Fuente2024) and cows (Wilms et al., Reference Wilms, Hare, Fischer-Tlustos, Vahmani, Dugan, Leal and Steele2022), C16:0 was the predominant FA in cow colostrum and transition milk, followed by C14:0 and C18:0. As an important component of milk fatty acids, production of BCFA are closely associated with the production of branched-chain fatty acids such as valine, leucine and isoleucine (Kaneda, Reference Kaneda1991). Similar to our findings, Xin et al. (Reference Xin, Xu, Chen, Chen, Steele and Guan2020) reported higher concentrations of total BCFA, iso-BCFA, anteiso-BCFA in transition milk (fifth milking) than that in colostrum (first milking) of both primiparous and multiparous dairy cows. However, limited knowledge is available regarding the profile and abundance of iso- and anteiso-BCFA in colostrum and transition milk especially from dairy goats. Fan et al. (Reference Fan, Zhang, Zhao, Ji, Du, Han, Wang and Yang2023) reported a numerically higher concentration of total BCFA in transition milk (1.38 g/100 g; collected at 3 d) than that in colostrum (1.31 g/100 g; collected at 1 d) of Laoshan dairy goats. Considering that the generation of BCFA is highly associated with factors including breed, dietary strategy and physiological conditions of cows (Bainbridge et al., Reference Bainbridge, Cersosimo, Wright and Kraft2016; Liu et al., Reference Liu, Wang, Guo, Huo, Zhang, Pei, Zhang and Wang2018; Roman-Garcia, Reference Roman-Garcia2019), further studies are needed to explore the profile and abundance of BCFA collected from dairy goats using a larger sample size.
The observed associations between certain bacterial genera and long-chain saturated FA indicate possible metabolic interactions within the milk ecosystem. Positive correlations between Fusobacterium and iso-C15:0 or anteiso-C17:0 and between Streptococcus and anteiso-C17:0 indicate that these bacteria either tolerate or selectively metabolise BCFA. The BCFA are known to improve membrane fluidity in bacteria, confer antimicrobial properties and support gut health through anti-inflammatory effects (Ran-RessleRan-Ressler R et al., Reference Ran-Ressler R, Khailova, Arganbright, Adkins-Rieck, Jouni, Koren, Ley, Brenna and Dvorak2011). These interactions may reflect selection pressures within milk that shape microbial communities while enhancing the bioactive profile of milk. The presence of BCFA in association with responsive bacterial taxa suggests a functional interdependence that may benefit both microbial viability and host physiology. These findings support the emerging view that milk is a biologically active matrix, containing immune, metabolic and microbiome signals that together influence neonatal health. Understanding these interactions offers new opportunities for nutritional interventions, such as modulation of maternal nutrition or microbial enrichment, to optimise milk quality and offspring development.
Despite the current findings, it should be noted that a relatively small sample size was studied in our experiment. In addition, all of the goats we used were from a single herd, which could be potential limitations of this study. In this regard, the bacterial composition identified in our study should be interpreted with caution when compared with the results obtained from other dairy goat breeds. Future study using a large amount of samples are needed to validate our findings and to reveal if the effect of breed, farm or parity could have impact on the profiles and abundance of bacteria in colostrum and transition milk of dairy goats.
Conclusion
This study describes the compositional and functional transitions between CM and TM in Guanzhong dairy goats, reflecting a shift from immunological to metabolic support of the neonate. A marked decrease in immunoglobulin concentrations in the TM was accompanied by an increase in long-chain saturated FA, BCFA and specific FA, indicating improved energy supply. The TM lipid profile, dominated by C16:0 and C18:0, is consistent with the increasing metabolic needs of early growth, while increased BCFAs may provide additional benefits through microbial modulation and anti-inflammatory activity. Despite a conserved core microbiota, taxonomic and functional changes revealed a stage-specific microbial role, with Emticicia enriched in the CM and Faecalibacterium and Ruminococcus enriched in the TM. Functional predictions emphasised stress resistance and heme biosynthesis in the CM and carbohydrate degradation pathways in the TM. The identification of Sphingopyxis, a genus capable of degrading xenobiotics, suggests a previously unrecognised role in milk quality. These results emphasise the dynamic interaction between milk composition, microbial ecology and neonatal development. One of the limitations of the current study is that only a small sample size was used and all dairy goats were selected from a single herd. The effect of breed, farm or parity on the profiles and abundance of bacteria and concentrations of BCFA in colostrum and transition milk of dairy goats needs to be verified using a larger sample size in future studies.
Acknowledgements
This research was supported by the China Agriculture Research System of MOF and MARA and Agricultural Sciences and Agricultural Science and Technology Innovation Program (CAAS-ASTIP-2022-FRI-04). We appreciate Dr. Qinan Zhao and Ms. Ziqi Meng from Inner Mongolia Academy of Agricultural & Animal Husbandry Sciences for their help in sample collection.
