Artificial Intelligence (AI) is reshaping the world as we know it, impacting all aspects of modern society, basically due to the advances in computer power, data availability and AI algorithms. The dairy sector is also on the move, from the exponential growth in AI research, to ready to use AI-based products, this new evolution to Dairy 4.0 represents a potential ‘game-changer’ for the dairy sector, to confront challenges regarding sustainability, welfare, and profitability. This research reflection explores the possible impact of AI, discusses the main drivers in the field and describes its origins, challenges, and opportunities. Further, we present a multidimensional vision considering factors that are not commonly considered in dairy research, such as geopolitical aspects and legal regulations that can have an impact on the application of AI in the dairy sector. This is just the beginning of the third tide of AI, and a future is still ahead. For now, the current advances in AI at on-farm level seem limited and based on the revised data, we believe that AI can be a ‘game-changer’ only if it is integrated with other components of Dairy 4.0 (such as robotics) and is fully adopted by dairy farmers.

The aim of this study was to examine the suitability of different growth functions (linear, sinusoidal, Gompertz, Schumacher and Richards) to fit cumulative milk production data from buffalo cows. Cumulative milk production at each day in milk was calculated from two published datasets reporting (i) fortnightly test-day milk yield records of the first lactation of Murrah buffalo that had calved during 1977–2012 and (ii) the first lactation records of Jaffarabadi buffalo collected from history-cum-pedigree registers for each quinquennium between 1991 and 2010. Each function was fitted to the lactation curves using nonlinear regression procedures. The Richards and sinusoidal equations provided the smallest root mean square error values, Akaike's and Bayesian information criteria and, therefore, the best fit for the cumulative lactation curves for milk yield. The Richards equation appeared to provide the most accurate estimate of the cumulative milk production at peak milk yield. Sinusoidal and flexible classical growth functions are appropriate to describe cumulative milk production curves and estimate lactation traits in buffalo.

This review describes the potentials of a new omics science in the dairy sector, particularly regarding the improvement of animal health and welfare. The three-dimensional network of water hydrogen bonds is a dynamic entity, subject to the influence of its components and properties of the environment. For this reason, it is sensitive to any chemical, physical or biological perturbation of the system. Therefore, the aqueous matrix acts as a sensitive sensor, reflecting the state, behaviour, and functionality of the system and providing similar information on its components. Aquaphotomics builds upon these underlying assumptions. It is a scientific discipline that combines spectroscopy with multivariate spectral analysis to extract the absorbance spectral pattern of water, which describes how water molecular structure changes in response to perturbations. Many studies assessed the applicability of this approach, including veterinary diagnostics. The water spectral pattern can be used as a multidimensional biomarker for rapid and non-invasive discrimination between healthy and diseased systems, even in the subclinical phase. Adopting such an approach, focused on precision farming, can foster subsequent optimisation of animal production performance and improve the overall profitability of farming operations.

Bayesian multiple regression models are often used for genomic selection, where all markers are adjusted simultaneously as random effects to reduce the false-discovery rate. The purpose of this research was to identify individual candidate genes for some reproductive traits in Holstein cattle through genome wide association studies (GWAS) using a Bayesian method in combination with post-GWAS analysis. Reproductive traits included: days open (DO), pregnancy rate (PR), calving interval (CI) and age at first calving (AFC). The animals were genotyped using single-nucleotide polymorphism (SNP) panels of different densities imputed to a 50 K SNP density. After quality control, we included 2400 genotyped animals. According to the Bayesian analysis, there were 19 windows with an explained additive genetic variance of >0.1 percent for CI, DO, AFC and PR in Holstein cattle, which were 3, 3, 6 and 7, respectively. Using Bayesian analysis, 79 genes were located within or nearby (250-kb) 19 significant SNPs/windows in the Bos taurus autosomes. Among these genes, we identified 25 candidate genes for reproductive traits, namely CHD7, CLVS1, EVX2, MAT2B, NUDCD2, GPR39, NCKAP5, LYPD1, HOXD13, SEMA5B, CCNG1, SEMA5A, BRF1, PSEN2, CACHD1, SUGTA, ELF1, SNORA70, AKT1, TM2D1, SLF1, MCTPA, PAB2A, MTRF1 and ADCY2. Additionally, another nine candidate genes (CLVS1, GPR39, CENPF, AMOT, ARF1, CCDC186, ADCY2, BDP1 and AMOTL1) were identified in the network cluster analysis as hub genes for reproductive traits. The results of gene set enrichment analysis (GSEA) and pathway analysis, suggest that the most important gene ontology term involving cellular metabolic process was related to the AFC trait. To summarize, Bayesian methods were used to identify SNPs and candidate genes that could be useful in genomic selection to improve reproductive traits of Holstein dairy cattle.

Signal transducers and activators of transcription (STAT) genes are involved in signal mediation of various hormones and cytokines. STAT1 located on chromosome number 2 is involved in mammary gland development and is associated with milk composition traits in bovines. This study aimed to find any relationship and impact of STAT1/BspHI gene with milk fat and protein yields in a herd of Holstein Friesian (HF) crossbred cattle of sub-tropical climate of Northern India. Milk composition data of 535 adult HF crossbred cows for a period of 12 years was collected from the records maintained at Livestock Farm, Guru Angad Dev Veterinary and Animal Sciences University. First lactation data of 222 animals was chosen for further analysis. After data correction for non-genetic factors (season of calving, period of calving, interaction effect of season and period of calving and age at first calving) these animals were categorised into two groups based on corrected high and low milk fat and protein yields. Forty animals were then selected for blood collection and further laboratory analysis. Amplified using PCR-RFLP technique, the 314 bp STAT1 gene was digested using BspHI restriction enzyme. C-T polymorphism at nucleotide position 201 and 260 of the STAT1 amplicon was observed. At 201, for genotype AA and Aa, the genotypic frequencies were 0.80 and 0.20%. At 260, for genotype BB and Bb, the genotypic frequencies were 0.25 and 0.75%. Least square analysis showed a significant association of all genotypes with milk fat and protein yields. Hence, STAT1 can be used as a potential candidate gene to aid in better animal selection in breeding programmes.

This research paper addresses the hypothesis that there is an optimal amount of intestinally available oleic acid (provided via abomasal infusion) to produce higher-oleic acid milk fat with satisfactory functional characteristics of cream and butter oil. A control and four increasing doses of free fatty acids from high oleic sunflower oil (HOSFA) were infused into the abomasum of four lactating dairy cows in a crossover experimental design with 7-d periods. Treatments were: (1) control (no HOSFA infused), (2) HOSFA (250 g/d), (3) HOSFA (500 g/d), (4) HOSFA (750 g/d), and (5) HOSFA (1000 g/d). All treatments included meat solubles and Tween 80 as emulsifiers. Viscosity, overrun and whipping time as well as foam firmness and stability were evaluated in whipping creams (33% fat). Solid fat content (from 0 to 40°C), melting point and firmness were determined in butter oil. Whipping time of cream increased linearly and viscosity decreased linearly as infusion of HOSFA increased. Overrun displayed a quadratic response, decreasing when 500 g/d or more was infused. Foam firmness and stability were not affected significantly by HOSFA. For butter oil, melting point, firmness, and solid fat content decreased as HOSFA infusion increased. Changes in 21 TG fractions were statistically correlated to functional properties, with 6–10 fractions showing the highest correlations consistently. Decisions on the optimal amount of HOSFA were dependent on the dairy product to which milk fat is applied. For products handled at commercial refrigeration temperatures, such as whipping cream and butter oil, the 250 g/d level was the limit to maintain satisfactory functional qualities. Palmitic acid needed to be present in at least 20% in milk fat to keep the functional properties for the products.

The objective of this study was to determine the effect of dietary calcium soaps from garlic (Allium sativum) and willow (Salix babylonica) extracts on nematode loads, nutrient intake and digestibility, nitrogen balance and rumen fermentation kinetics in dairy goats. Nine adult non-lactating Saanen goats were grouped into a complete randomized block design with 3 treatments (n = 3) over a period of 28 d. Animals were fed a diet based on alfalfa hay and a concentrate that was supplemented (65 g/kg DM) with calcium soaps of safflower (control), garlic or willow. Intake of dry matter (DM), organic matter (OM) and neutral detergent fiber (NDF) were not affected by dietary calcium soaps. However, the highest digestibility of DM and OM were observed in willow supplemented goats. In vitro gas kinetics and fermentation profile were not affected by diets. Results from fecal egg count indicated a reduction in total count, Haemonchus spp. and Trychostrongylus spp. for both garlic and willow compared to control. Our results suggest that calcium soaps of garlic or willow extracts can be used to reduce gastrointestinal parasites in goats without compromising productive traits or rumen function.

This research paper addresses the hypothesis that substituting soybean meal with locally produced yeast protein from Cyberlindnera jadinii in barley-based concentrates for Norwegian Red (NR) dairy cows does not have adverse effects on milk fatty acid (FA) composition, rumen microbiota and sensory quality of milk. As soybeans also represent valuable protein sources for human consumption, alternative protein sources need to be investigated for animal feed. A total of 48 NR dairy cows were allocated into three feeding treatments, with the same basal diet of grass silage, but different concentrates. The concentrates were all based on barley, but 7% of the barley in the barley-concentrate (BAR; negative control) was replaced by either soybean meal (SBM; conventional control) or yeast microbial protein (YEA). The experiment lasted for a total of 10 weeks, including 2 weeks of adaptation with the soybean meal concentrate. Analysis of the feed revealed some differences in the FA composition of the YEA concentrate compared to the SBM and BAR concentrates. In milk, only two FAs (C17:1n-8cis9 and an unidentified isomer of C18:3) were significantly different between the YEA- and SBM-group, while six FAs differed between the BAR- and SBM-group. However, the amount of these FAs was low compared to the entire FA profile (<0.7 g/100 g). The experimental diets did not affect rumen microbiota nor the milk sensory quality. This study shows that C. jadinii can replace soybean meal as a protein source in concentrates (7% inclusion) for NR dairy cows fed a diet composed of grass silage and concentrates without any effects on rumen microbiota, and without compromising the FA composition or sensory quality of milk.

This study aimed to estimate live body weight from body measurements for Holstein × Zebu dairy cows (n = 156) reared under conditions of humid tropics in Mexico using multivariate adaptive regression splines algorithm (MARS) with several train-test proportions. The body measurements included withers height, rump height, hip width, heart girth, body length and diagonal body length. The data were divided into 65:35, 70:30 and 80:20 split data for training and testing sets, respectively. The MARS algorithm was used to construct a prediction model, which predicted the body weight from the body measurements of the test dataset. The results emphasized that the MARS algorithm had an explanation rate for 80:20 train and test set of 0.836 and 0.711, respectively, with minimum Akaike information criterion values. This indicates that it is a reliable way of predicting body weight from body measurements. The results suggest that body weight prediction can be performed with the MARS algorithm in a reliable way, therefore, this algorithm may be a useful tool for animal breeders and researchers in the development of feeding and selection-aimed approaches.

The aim of this study was to evaluate parameters for the prediction of peripartum hypocalcemia in cattle. The study consisted of two independent investigations (A and B) conducted in different time periods and locations. Blood sera were analyzed for the activity of alkaline phosphatase (ALP), the concentration of ionized (iCa), and total calcium (Ca) two weeks before calving. On the day of calving, urinary net acid/base excretion was calculated, and the serum Ca concentration was analyzed. Ca concentration less than 2 mmol/l on the day of calving was defined as hypocalcemia. The results differed between the two investigations. Part A showed high predictive values for the antepartum parameters ALP and Ca concentration, but these could not be replicated in part B. Animals enrolled were quite young (average age 5.5 years) and the group of animals with hypocalcemia was comparably small in both investigations. The reason(s) for the different results in investigation A and B are not clear. These findings indicate that there are probably as yet unidentified parameters and influences on calcium homeostasis in cattle. None of the parameters investigated in the present study can be considered a reliable universal parameter for prediction of hypocalcemia risk in general.

The work reported in the Research Communication investigated in vitro rumen gas kinetics and fermentation profile as well as in vivo performance of lactating ewes fed corn silage (CS), sunflower silage (SFS) and their 50 : 50 mixture (CS-SFS). For the in vivo experiment, nine early-lactation Suffolk × Texel ewes were grouped in a replicated 3 × 3 Latin square design of three 21-d periods. Treatments were based on ad libitum CS, SFS, and CS-SFS supplemented with concentrate at 48 g/kg LW0.75. In vitro results showed that the CS had the highest dry matter degraded substrate and microbial crude protein production followed by CS-SFS. The in vivo data showed that animals fed on CS had higher digestibility of dry matter and organic matter than CS-SFS, while SFS were intermediate. Nitrogen (N) intake, fecal N excretion, and urine N excretion were similar between groups, however, milk N excretion was lower in SFS than CS. Milk yield was higher for CS and CS-SFS than SFS group, however, SFS-fed ewes had higher milk fat content than either CS or CS-SFS (all differences reported here were significant, P < 0.05 or better). Overall, CS-SFS could be used as dietary roughage for dairy ewes without deleterious effects on nutrient intake, N-balance and milk yield whilst potentially offering a more sustainable alternative to CS.

The aim of the study was to evaluate the effect of total mixed ration particle size (length) and breed of cow on intake dynamics, animal performance and CH4 emissions, comparing high yielding Holstein and low yielding Girolando cows. The experimental design was 2 × 2 Latin Square arranged as a crossover factorial scheme with two diets (short particle size, SPS and long particle size, LPS) and the two breed compositions. The design comprised two periods of 26 d each, where all data collection was performed at cow level. No influence of the particle size occurred for the passage rate, neutral detergent fiber digestibility, performance and milk composition, methane emissions or ruminal fermentation parameters. Girolando cows had greater dry matter intake (DMI) when fed SPS, while Holsteins had the same (P < 0.05). Girolando cows had lower dry matter digestibility when fed LPS compared to SPS, while Holsteins had the opposite effect (P < 0.05). Also, the digestibility of crude protein and non-fibrous carbohydrates decreased in Girolando cows fed LPS, but not in Holsteins (P < 0.05). Girolando cows reduced DMI by 10.6% when fed LPS diet (P < 0.05). Girolando had an increased eating rate (+24 g of DM/min; P < 0.05) compared to Holstein cows, but Holstein cows had a lower CH4 intensity (by 29.7%: P < 0.05). Girolando cows increased the dry matter intake when fed a diet with short particle size, while the same did not happen in Holsteins. Dry matter digestibility increased in Holsteins when fed long particle size, while the opposite was observed in Girolando cows. Nutrient digestibility was reduced in Girolando cows when fed short particle size. Particle size did not influence eating time, eating rate, feed trough visits, visits with intake, milk yield and composition regardless of the breed. Reducing particle size increased CH4 intensity in both breeds.

The aim of this experiment was to investigate the differential proteomic characteristics of milk from high- and low-yielding Guanzhong dairy goats during the peak lactation period under the same feeding conditions. Nine Guanzhong dairy goats with high yield (H: 3.5 ± 0.17 kg/d) and nine with low yield (L:1.2 ± 0.25 kg/d) were selected for milk proteomic analysis using tandem mass tag technology. A total of 78 differentially expressed proteins were identified. Compared with L, 50 proteins including HK3, HSPB1 and ANXA2 were significantly upregulated in H milk, while 28 proteins including LALBA and XDH were significantly downregulated. Bioinformatics analysis of the differentially expressed proteins showed that galactose metabolism, purine metabolism, glycolysis/gluconeogenesis, MAPK signaling pathway, regulation of actin cytoskeleton and other pathways were closely related to milk yield. HK3, HSPB1, ANXA2, LALBA and XDH were important candidate proteins associated with the milk production characteristics of Guanzhong dairy goats. Our data provide relevant biomarkers and a theoretical basis for improving milk production in Guanzhong dairy goats.

This research paper proposes a simple image processing technique for automatic lameness detection in dairy cows under farm conditions. Seventy-five cows were selected from a dairy farm and visually assessed for a reference/real lameness score (RLS) as they left the milking parlor, while simultaneously being video-captured. The method employed a designated walking path and video recordings processed through image analysis to derive a new computerized automatic lameness score (ALDS) based on calculated factors from back arch posture. The proposed automatic lameness detection system was calibrated using 12 cows, and the remaining 63 were used to evaluate the diagnostic characteristics of the ALDS. The agreement and correlation between ALDS and RLS were investigated. ALDS demonstrated high diagnostic accuracy with 100% sensitivity and specificity and was found to be 100% accurate with a perfect agreement (ρc = 1) and strong correlation (r = 1, P < 0.001) for lameness detection in binary scores (lame/non-lame). Moreover, the ALDS had a strong agreement (ρc = 0.885) and was highly correlated (r = 0.840; 0.796–1.000 95% confidence interval, P < 0.001) with RLS in ordinal scores (lameness severity; LS1 to LS5). Our findings suggest that the proposed method has the potential to compete with vision-based lameness detection methods in dairy cows in farm conditions.

This research communication reports the effects of a compound enzyme preparation consisting of fibrolytic (cellulase 3500 CU/g, xylanase 2000 XU/g, β-glucanase 17 500 GU/g) and amylolytic (amylase 37 000 AU/g) enzymes on nutrient intake, rumen fermentation, serum parameters and production performance in primiparous early-lactation (47 ± 2 d) dairy cows. Twenty Holstein–Friesian cows in similar body condition scores were randomly divided into control (CON, n = 10) and experimental (EXP, n = 10) groups in a completely randomized single-factor design. CON was fed a basal total mixed ration diet and EXP was dietary supplemented with compound enzyme preparation at 70 g/cow/d. The experiment lasted 4 weeks, with 3 weeks for adaptation and then 1 week for measurement. Enzyme supplementation significantly increased diet non-fibrous carbohydrates (NFC) content as well as dry matter intake (DMI) and NFC intake (P < 0.05). EXP had increased ruminal butyrate and isobutyrate percentages (P < 0.01) but decreased propionate and valerate percentages (P < 0.05), as well as increased serum alkaline phosphatase activity and albumin concentration (P ≤ 0.01). Additionally, EXP had increased milk yield (0.97 kg/d), 4% fat corrected milk yield and energy corrected milk yield, as well as milk fat and protein yield (P < 0.01). In conclusion, dietary supplementation with a fibrolytic and amylolytic compound enzyme preparation increased diet NFC content, DMI and NFC intake, affected rumen fermentation by increasing butyrate proportion at the expense of propionate, and enhanced milk performance in primiparous early-lactation dairy cows.

This research paper describes a validation study evaluating the ability of IceTag accelerometers (Peacock Technology, UK) to detect play behaviour in weaned dairy calves. Play behaviour is commonly observed in young animals and is regarded as an indicator of positive welfare states. Eight Holstein Friesian calves aged three to five months old were monitored using leg-mounted accelerometers for 48 h. Data generated by accelerometers to quantify calf activity included step count, lying times and a proprietary measure of overall activity termed ‘motion index’ (MI). Calf behaviour was filmed continuously over the same 48-h period using closed circuit television cameras and analysed using one-zero sampling to identify the presence (1) or absence (0) of play within each 15-min time period. A positive correlation between MI and visually recorded play was found. Visual observations were compared with accelerometer-generated data and analysed using 2 × 2 contingency tables and classification and regression tree analysis. A MI value of ≥69 was established as the optimum threshold to detect play behaviour (sensitivity = 94.4%; specificity = 93.6%; balanced accuracy = 94.0%). The results of this study suggest that accelerometer-generated MI data have the potential to detect play behaviour in weaned dairy calves in a more time efficient manner than traditional visual observations.

Minerals are supplemented routinely to dairy cows during the dry period to prevent metabolic issues postpartum. However, limited information exists on the impacts of mineral supplementation on colostrum carotenoids. This study aimed to determine the effects of prepartum supplementation with three micro-nutrients; inorganic selenium (INORG), organic selenium (ORG) or rumen-protected choline (RPC) on the carotenoid content of bovine colostrum and transition milk (TM) from pasture-based dairy cows. A total of 57 (12 primiparous and 45 multiparous) Holstein-Friesian (HF) and HF × Jersey (JEX) cows were supplemented daily for 49 ± 12.9 d before calving. Colostrum samples were collected from all cows immediately postpartum and TM one to five (TM1–TM5) were collected from a sub-set of 15 cows (five per treatment group) at each consecutive milking postpartum. Carotenoid concentration was determined using ultra-high performance liquid chromatography – diode array detection (UHPLC-DAD). With the use of transmittance, the colour index and colour parameters a*, b* and L* were used to determine colour variations over this period. Prepartum supplementation did not have a significant effect on colostrum β-carotene concentration or colour. Positive correlations between β-carotene and colour parameter b* (R2 = 0.671; P < 0.001) and β-carotene and colour index (R2 = 0.560; P < 0.001) were observed. Concentrations of β-carotene were highest in colostrum (1.34 μg/g) and decreased significantly with each milking postpartum (TM5 0.31 μg/g). Breed had a significant effect on colostrum colour with JEX animals producing a greater b* colostrum than HF animals (P = 0.030). Primiparous animals produced colostrum with the weakest colour compared to second or ≥third parity animals (P = 0.042). Despite statistical increases in the b* parameter in colostrum from JEX cows and multiparous cows, β-carotene concentrations did not significantly increase suggesting that other factors may influence colostrum colour. The b* parameter may be used as an indicator for estimating carotenoid concentrations in colostrum and TM, particularly when assessed via transmittance spectroscopy.