The aim of this study was to investigate associations between pathogen-specific cases of subclinical mastitis and milk yield, quality, protein composition, and cheese-making traits. Forty-one multibreed herds were selected for the study, and composite milk samples were collected from 1,508 cows belonging to 3 specialized dairy breeds (Holstein Friesian, Brown Swiss, and Jersey) and 3 dual-purpose breeds of Alpine origin (Simmental, Rendena, and Grey Alpine). Milk composition [i.e., fat, protein, casein, lactose, pH, urea, and somatic cell count (SCC)] was analyzed, and separation of protein fractions was performed by reversed-phase high performance liquid chromatography. Eleven coagulation traits were measured: 5 traditional milk coagulation properties [time from rennet addition to milk gelation (RCT, min), curd-firming rate as the time to a curd firmness (CF) of 20 mm (k20, min), and CF at 30, 45, and 60 min from rennet addition (a30, a45, and a60, mm)], and 6 new curd firming and syneresis traits [potential asymptotical CF at an infinite time (CFP, mm), curd-firming instant rate constant (kCF, % × min-1), curd syneresis instant rate constant (kSR, % × min-1), modeled RCT (RCTeq, min), maximum CF value (CFmax, mm), and time at CFmax (tmax, min)]. We also measured 3 cheese yield traits, expressing the weights of total fresh curd (%CYCURD), dry matter (%CYSOLIDS), and water (%CYWATER) in the curd as percentages of the weight of the processed milk, and 4 nutrient recovery traits (RECPROTEIN, RECFAT, RECSOLIDS, and RECENERGY), representing the percentage ratio between each nutrient in the curd and milk. Milk samples with SCC > 100,000 cells/mL were subjected to bacteriological examination. All samples were divided into 7 clusters of udder health (UH) status: healthy (cows with milk SCC < 100,000 cells/mL and uncultured); culture-negative samples with low, medium, or high SCC; and culture-positive samples divided into contagious, environmental, and opportunistic intramammary infection (IMI). Data were analyzed using a linear mixed model. Significant variations in the casein to protein ratio and lactose content were observed in all culture-positive samples and in culture-negative samples with medium to high SCC compared to normal milk. No differences were observed among contagious, environmental, and opportunistic pathogens, suggesting an effect of inflammation rather than infection. The greatest impairment in milk quantity and composition, clotting ability, and cheese production was observed in the 2 UH status groups with the highest milk SCC (i.e., contagious IMI and culture-negative samples with high SCC), revealing a discrepancy between the bacteriological results and inflammatory status, and thus confirming the importance of SCC as an indicator of udder health and milk quality.

Cheese yield (CY) is the most important technological trait of milk, because cheese-making uses a very high proportion of the milk produced worldwide. Few studies have been carried out at the level of individual milk-producing animals due to a scarcity of appropriate procedures for model-cheese production, the complexity of cheese-making, and the frequent use of the fat and protein (or casein) contents of milk as a proxy for cheese yield. Here, we report a high-throughput cheese manufacturing process that mimics all phases of cheese-making, uses 1.5-L samples of milk from individual animals, and allows the simultaneous processing of 15 samples per run. Milk samples were heated (35°C for 40 min), inoculated with starter culture (90 min), mixed with rennet (51.2 international milk-clotting units/L of milk), and recorded for gelation time. Curds were cut twice (10 and 15 min after gelation), separated from the whey, drained (for 30 min), pressed (3 times, 20 min each, with the wheel turned each time), salted in brine (for 60 min), weighed, and sampled. Whey was collected, weighed, and sampled. Milk, curd, and whey samples were analyzed for pH, total solids, fat content, and protein content, and energy content was estimated. Three measures of percentage cheese yield (%CY) were calculated: %CY(CURD), %CY(SOLIDS), and %CY(WATER), representing the ratios between the weight of fresh curd, the total solids of the curd, and the water content of the curd, respectively, and the weight of the milk processed. In addition, 3 measures of daily cheese yield (dCY, kg/d) were defined, considering the daily milk yield. Three measures of nutrient recovery (REC) were computed: REC(FAT), REC(PROTEIN), and REC(SOLIDS), which represented the ratio between the weights of the fat, protein, and total solids in the curd, respectively, and the corresponding components in the milk. Energy recovery, REC(ENERGY), represented the energy content of the cheese compared with that in the milk. This procedure was used to process individual milk samples obtained from 1,167 Brown Swiss cows reared in 85 herds of the province of Trento (Italy). The assessed traits exhibited almost normal distributions, with the exception of REC(FAT). The average values (± SD) were as follows: %CY(CURD)=14.97±1.86, %CY(SOLIDS)=7.18±0.92, %CY(WATER)=7.77±1.27, dCY(CURD)=3.63±1.17, dCY(SOLIDS)=1.74±0.57, dCY(WATER)=1.88±0.63, REC(FAT)=89.79±3.55, REC(PROTEIN)=78.08±2.43, REC(SOLIDS)=51.88±3.52, and REC(ENERGY)=67.19±3.29. All traits were highly influenced by herd-test-date and days in milk of the cow, moderately influenced by parity, and weakly influenced by the utilized vat. Both %CY(CURD) and dCY(CURD) depended not only on the fat and protein (casein) contents of the milk, but also on their proportions retained in the curd; the water trapped in curd presented an higher variability than that of %CY(SOLIDS). All REC traits were variable and affected by days in milk and parity of the cows. The described model cheese-making procedure and the results obtained provided new insight into the phenotypic variation of cheese yield and recovery traits at the individual level.

The aims of this study were to investigate variation of milk coagulation property (MCP) measures and their predictions obtained by mid-infrared spectroscopy (MIR), to investigate the genetic relationship between measures of MCP and MIR predictions, and to estimate the expected response from a breeding program focusing on the enhancement of MCP using MIR predictions as indicator traits. Individual milk samples were collected from 1,200 Brown Swiss cows (progeny of 50 artificial insemination sires) reared in 30 herds located in northern Italy. Rennet coagulation time (RCT, min) and curd firmness (a(30), mm) were measured using a computerized renneting meter. The MIR data were recorded over the spectral range of 4,000 to 900 cm(-1). Prediction models for RCT and a(30) based on MIR spectra were developed using partial least squares regression. A cross-validation procedure was carried out. The procedure involved the partition of available data into 2 subsets: a calibration subset and a test subset. The calibration subset was used to develop a calibration equation able to predict individual MCP phenotypes using MIR spectra. The test subset was used to validate the calibration equation and to estimate heritabilities and genetic correlations for measured MCP and their predictions obtained from MIR spectra and the calibration equation. Point estimates of heritability ranged from 0.30 to 0.34 and from 0.22 to 0.24 for RCT and a(30), respectively. Heritability estimates for MCP predictions were larger than those obtained for measured MCP. Estimated genetic correlations between measures and predictions of RCT were very high and ranged from 0.91 to 0.96. Estimates of the genetic correlation between measures and predictions of a(30) were large and ranged from 0.71 to 0.87. Predictions of MCP provided by MIR techniques can be proposed as indicator traits for the genetic enhancement of MCP. The expected response of RCT and a(30) ensured by the selection using MIR predictions as indicator traits was equal to or slightly less than the response achievable through a single measurement of these traits. Breeding strategies for the enhancement of MCP based on MIR predictions as indicator traits could be easily and immediately implemented for dairy cattle populations where routine acquisition of spectra from individual milk samples is already performed.

Fourier-transform infrared (FTIR) spectra are used to predict the fat, protein, casein, and lactose contents of milk. These estimates are currently used to predict the individual estimated breeding values of animals. The objective of the present study was to estimate the genetic variation and heritabilities of the milk transmittance spectrum at each individual FTIR wave. Milk was sampled once per cow from a total of 1,064 Italian Brown Swiss cows from 30 herds, sired by 50 artificial insemination sires. The FTIR spectra of all samples were collected within 3 h of sampling from 25 mL of milk. The obtained spectral range comprised wavenumbers 5,000 to 930×cm(-1), corresponding to wavelengths 2.00 to 10.76 μm and frequencies from 149.9 to 27.9 THz, for a total of 1,056 waves. These were acquired using a MilkoScan FT120 FTIR interferometer (Foss Electric A/S, Hillerød, Denmark). Each spectral data point was treated as a single trait and analyzed using an animal model REML method. The results indicated that the transmittance of the bovine milk FTIR spectrum was heritable for most individual waves in the wavenumber interval from 5,000 to 930×cm(-1). Moreover, the transmittance of contiguous FTIR waves was much more highly correlated in terms of the average value and phenotypic variation, compared with genetic variation. In the present study, we characterized 5 regions of the FTIR spectrum that were relevant to the analysis of milk; 2 regions, one in the transition area between the short-wavelength infrared (SWIR) and mid-wavelength infrared (MWIR) divisions of the electromagnetic spectrum (SWIR-MWIR region) and another very short region in the MWIR division (MWIR-2 region), were characterized by very high phenotypic variability in the transmittance of individual milk samples within each wave. This was caused by the absorption peaks of water, which can mask the effects of other important milk components. These regions also showed high genetic variability in transmittance, and the heritability estimates of individual waves were generally very low (with some exceptions). The 3 other identified regions contained many transmittance peaks that represented important chemical bonds; these showed much lower phenotypic and genetic variability in terms of individual waves, but relatively higher and less variable heritability estimates. Among them, the SWIR region (near-infrared) showed a peculiar cyclic pattern of the heritability coefficients of transmittance, the MWIR-1 region was particularly important for the estimation of fat, and the MWIR-LWIR region (also known also as the "fingerprint region") had 3 areas of relatively high heritability. In summary, we found that the transmittance data from the FTIR spectra of milk have genetic variability that may prove useful for the direct genetic improvement of dairy species, rather than only through indirect phenotypic predictions of individual milk quality and technological traits.

The aim of this study was to assess the performance of Bayesian models commonly used for genomic selection to predict "difficult-to-predict" dairy traits, such as milk fatty acid (FA) expressed as percentage of total fatty acids, and technological properties, such as fresh cheese yield and protein recovery, using Fourier-transform infrared (FTIR) spectral data. Our main hypothesis was that Bayesian models that can estimate shrinkage and perform variable selection may improve our ability to predict FA traits and technological traits above and beyond what can be achieved using the current calibration models (e.g., partial least squares, PLS). To this end, we assessed a series of Bayesian methods and compared their prediction performance with that of PLS. The comparison between models was done using the same sets of data (i.e., same samples, same variability, same spectral treatment) for each trait. Data consisted of 1,264 individual milk samples collected from Brown Swiss cows for which gas chromatographic FA composition, milk coagulation properties, and cheese-yield traits were available. For each sample, 2 spectra in the infrared region from 5,011 to 925 cm(-1) were available and averaged before data analysis. Three Bayesian models: Bayesian ridge regression (Bayes RR), Bayes A, and Bayes B, and 2 reference models: PLS and modified PLS (MPLS) procedures, were used to calibrate equations for each of the traits. The Bayesian models used were implemented in the R package BGLR (http://cran.r-project.org/web/packages/BGLR/index.html), whereas the PLS and MPLS were those implemented in the WinISI II software (Infrasoft International LLC, State College, PA). Prediction accuracy was estimated for each trait and model using 25 replicates of a training-testing validation procedure. Compared with PLS, which is currently the most widely used calibration method, MPLS and the 3 Bayesian methods showed significantly greater prediction accuracy. Accuracy increased in moving from calibration to external validation methods, and in moving from PLS and MPLS to Bayesian methods, particularly Bayes A and Bayes B. The maximum R(2) value of validation was obtained with Bayes B and Bayes A. For the FA, C10:0 (% of each FA on total FA basis) had the highest R(2) (0.75, achieved with Bayes A and Bayes B), and among the technological traits, fresh cheese yield R(2) of 0.82 (achieved with Bayes B). These 2 methods have proven to be useful instruments in shrinking and selecting very informative wavelengths and inferring the structure and functions of the analyzed traits. We conclude that Bayesian models are powerful tools for deriving calibration equations, and, importantly, these equations can be easily developed using existing open-source software. As part of our study, we provide scripts based on the open source R software BGLR, which can be used to train customized prediction equations for other traits or populations.

Blood serum proteins are significant indicators of animal health. Nevertheless, several factors should be considered to appropriately interpret their concentrations in blood. Therefore, the objectives of this study were (1) to assess the effect of herd productivity, breed, age and stage of lactation on serum proteins and (2) to investigate association between serum proteins and somatic cell count (SCC) in dairy cattle. Milk and blood samples were collected from 1508 cows of six different breeds (Holstein Friesian, Brown Swiss, Jersey, Simmental, Rendena and Alpine Grey) that were housed in 41 multi-breed herds. Milk samples were analyzed for composition and SCC, while blood samples were analyzed for serum proteins (i.e. total protein, albumin, globulin and albumin-to-globulin ratio (A : G)). Herds were classified as low or high production, according to the cow’s average daily milk energy yield adjusted for breed, days in milk (DIM) and parity. Data were analyzed using a linear mixed model that included the fixed effects of DIM, parity, SCS, breed, herd productivity and the random effect of the Herd-test date within productivity level. Cows in high producing herds (characterized also by greater use of concentrates in the diet) had greater serum albumin concentrations. Breed differences were reported for all traits, highlighting a possible genetic mechanism. The specialized breed Jersey and the two dual-purpose local breeds (Alpine Grey and Rendena) had the lowest globulin concentration and greatest A : G. Changes in serum proteins were observed through lactation. Total protein reached the highest concentration during the 4th month of lactation. Blood albumin increased with DIM following a quadratic pattern, while globulin decreased linearly. As a consequence, A : G increased linearly during lactation. Older cows had greater total protein and globulin concentrations, while albumin concentration seemed to be not particularly affected by age. A linear relationship between serum proteins and SCS was observed. High milk SCS was associated with greater total protein and globulin concentrations in blood. The rise in globulin concentration, together with a decrease in albumin concentrations, resulted in a decline in A : G as SCS of milk increased. In conclusion, such non-genetic factors must be considered to appropriately interpret serum proteins as potential animal welfare indicator and their evaluation represents an important first-step for future analysis based on the integration of metabolomics, genetic and genomic information for improving the robustness of dairy cows.