This article focuses on the creation of a monitoring tool using routinely collected data from milk payment analyses. Milk samples were analyzed through Fourier Transform mid-infrared spectrometry every 1 to 3 days, and their compositions were predicted using machine learning models. Among the predicted parameters, fatty acid profiles appear to be effective indicators of animal status and management practices. In this research, these profiles were summarized using 31 fatty acids or groups of fatty acids. The methodology consists of four steps: hierarchical clustering to detect patterns in a Belgian spectral dataset (N = 774,781), interpretation of the identified seven clusters, development of predictive models applied to a Canadian dataset (N = 670,165), and validation using management information collected from Canadian farms. The identified clusters revealed significant relationships with feeding management strategies and temporal evolutions, highlighting the potential to develop automated alert systems that assist farmers and advisors in herd monitoring.

Although improving the management of lactating cows to reduce health and reproductive issues can enhance cow longevity, the long-term effects of early-life management practices are less understood. The objectives of this study were to characterize dairy farms based on their early-life management practices and analyze their associations with herd longevity, productivity, and profitability. In this cross-sectional observational study, early-life management practices regarding colostrum feeding, milk feeding, solid feed and weaning, and housing were collected from 1,658 dairy farms in Québec, Canada, using a questionnaire between February 2020 and February 2021. Length of productive life and the percentage of cows in their third or greater lactation, estimated from DHI testing data, were used as herd longevity indicators, whereas lifetime cumulative ECM production and lifetime cumulative milk value, also derived from DHI records, served as indicators of productivity and profitability, respectively. Cluster analysis was performed to characterize farms based on their early-life management practices. Cluster stability assessment was used to determine the best clustering algorithm and number of clusters. Associations between herd longevity, productivity, profitability, and early-life management practices were assessed using multivariate linear regression models. Due to missing data (ranging from 0.1% to 15.4% across variables), multiple imputation was employed, and significant practices were identified by iteratively applying likelihood ratio tests (α < 0.05) across the imputed datasets. Two clusters were identified and denominated as traditionally or modernly managed farms. The traditionally managed farms cluster (n = 600; 36.2%) was characterized by feeding nonpasteurized or nonacidified milk (whole or waste) with individual buckets, not measuring the concentration of IgG in the colostrum, and housing calves individually. Modernly managed farms (n = 1,058; 63.8%) were characterized by feeding calves powdered milk replacer through automated systems and group housing calves both before and after weaning. Practices adopted by traditionally managed farms were associated with increased longevity but lower productivity and profitability, whereas practices adopted by modernly managed farms were associated with lower longevity but increased productivity and profitability. Our results highlight that early-life management practices are linked with herd longevity, productivity, and profitability, but further research is needed to understand the underlying factors contributing to these associations and to guide dairy farmers in making informed management decisions.

The goal of this study was to isolate spectral fingerprints from milk Fourier transform infrared spectra that may reflect potential improvements in cow welfare, specifically comfort and ease of movement, resulting from modified housing configurations. Housing configuration modification treatments were tested across 3 animal trials, consisting of modified chain length (TCL), stall width (SW) and manger wall and stall length (MW/SL) configurations. The spectral analyses involved the use of principal components and mixed model analysis. Principal components were calculated from averages of mid-infrared spectra collected on the last weeks of treatment application in each of the animal trials. A significant effect of housing configuration was revealed. As an indication of animal comfort improvement, milk of cows assigned to longer chains revealed a trend of changes in multiple milk components (e.g., milk NPN, trans fatty acids, fat, and protein) that are consistent with changes in ruminal pH. These conclusions were inline with those drawn from the analysis of animal-based responses such as behavioral data and other outcomes. This study was able to reveal that housing modifications had a significant effect on milk spectra, with differences observed between the most and least restrictive treatments, translating into improved or reduced animal welfare status.

Our objective was to validate the possibility of detecting SARA from milk Fourier transform mid-infrared spectroscopy estimated fatty acids (FA) and machine learning. Subacute ruminal acidosis is a common condition in modern commercial dairy herds for which the diagnostic remains challenging due to its symptoms often being subtle, nonexclusive, and not immediately apparent. This observational study aimed at evaluating the possibility of predicting SARA by developing machine learning models to be applied to farm data and to provide an estimated portrait of SARA prevalence in commercial dairy herds. A first data set composed of 488 milk samples of 67 cows (initial DIM = 8.5 ± 6.18; mean ± SD) from 7 commercial dairy farms and their corresponding SARA classification (SARA+ if rumen pH <6.0 for 300 min, else SARA-) was used for the development of machine learning models. Three sets of predictive variables: i) milk major components (MMC), ii) milk FA (MFA), and iii) MMC combined with MFA (MMCFA) were submitted to 3 different algorithms, namely Elastic net (EN), Extreme gradient boosting (XGB), and Partial least squares (PLS), and evaluated using 3 different scenarios of cross-validation. Accuracy, sensitivity, and specificity of the resulting 27 models were analyzed using a linear mixed model. Model performance was not significantly affected by the choice of algorithm. Model performance was improved by including fatty acids estimations (MFA and MMCFA as opposed to MMC alone). Based on these results, one model was selected (algorithm: EN; predictive variables: MMCFA; 60.4, 65.4, and 55.3% of accuracy, sensitivity, and specificity, respectively) and applied to a large data set comprising the first test-day record (milk major components and FA within the first 70 DIM of 211,972 Holstein cows (219,503 samples) collected from 3001 commercial dairy herds. Based on this analysis, the within-herd SARA prevalence of commercial farms was estimated at 6.6 ± 5.29% ranging from 0 to 38.3%. A subsequent linear mixed model was built to investigate the herd-level factors associated to higher within-herd SARA prevalence. Milking system, proportion of primiparous cows, herd size and seasons were all herd-level factors affecting SARA prevalence. Furthermore, milk production was positively, and milk fat yield negatively associated with SARA prevalence. Due to their moderate levels of accuracy, the SARA prediction models developed in our study, using data from continuous pH measurements on commercial farms, are not suitable for diagnostic purpose. However, these models can provide valuable information at the herd level.

The use of prerecorded data to remotely assess the herd welfare status is a promising approach to reduce the need for costly and time-consuming on-farm welfare assessments. Therefore, the objective of this study was to validate the Herd Status Index, an index developed based on Dairy Herd Improvement data from Canada, to remotely evaluate the welfare status of dairy herds. Herd-level prevalence of five animal -based welfare outcomes, measured once on 2 986 Quebec - Canada dairy herds between 2016 and 2019, were used to generate clusters with different welfare status using the algorithm partitioning around medoids. Dairy Herd Improvement data from 12 months prior to the welfare assessment were extracted and used to calculate the Herd Status Index. A linear model was used to carry out comparisons between clusters. Three stable clusters were found to best describe the data. Cluster two had the best overall wel-fare status since it had the lowest prevalence of all welfare issues while cluster three had the highest prevalence of most welfare issues, with the exception for the prevalence of neck lesions that was not dif-ferent than cluster one. Cluster one had an overall intermediate welfare status. The Herd Status Index was higher (i.e., indicating a good welfare status) on cluster two compared to cluster three, but neither cluster three nor two differed to cluster one. In its current format, the Herd Status Index has a weak potential to identify herds with varying prevalence of welfare issues and it requires further improvements before it could be used to accurately assess the welfare status of the herds.(c) 2022 The Authors. Published by Elsevier B.V. on behalf of The Animal Consortium. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

Increasing the productive lifespan of dairy cows is important to achieve a sustainable dairy industry, but making strategic culling decisions based on cow profitability is challenging for farmers. The objective of this study was to carry out a lifetime cost-benefit analysis based on production and health records and to explore different culling decisions among farmers. The cost-benefit analysis was conducted for 22 747 dairy cows across 114 herds in Quebec, Canada for which feed costs and the occurrence of diseases were reported. Costs and revenues related to productive lifespan were compared among cohorts of cows that left their respective herd at the end of their last completed lactation or stayed for a complete additional lactation. Hierarchical clustering analysis was carried out based on costs and revenues to explore different culling decisions among farmers. Our results showed that the knowledge of lifetime cumulative costs and revenues was of great importance to identify low-profitable cows at an earlier lactation, while only focusing on current lactation costs and revenues can lead to an erroneous assessment of profitability. While culling decisions were mostly based on current lactation costs and revenues and disregarded the occurrence of costly events on previous lactations, there was variation among farmers as we identified three different culling decision clusters. Monitoring cumulative costs and revenues would help farmers to identify low-profitable cows at an earlier lactation and make the decision to increase herd productive lifespan and farm profitability by keeping the most profitable cows.

Nutritive value of a fodder from extensive established pasture was tested. The nutrient content was measured by the Wendeeanalysis and by in vitro ruminant digestibility method. Results of former experiments showed that the nutritive value of an extensive established pasture in the case of late outdoor growing is low. In our results the highest crude protein content was in the year 2002, while in 2003 can be observed a steep decline, which showed in 2004 further decrease. The crude protein values were the highest in case of middle seed norm. The nutritive values of these pastures provide just supply for the demand and it was declined due to the negative N-balance in the rumen. Our results showed that the samples from the year of establishment could possibly be used for preserved feed (6.01 MJ NE l kg-1). The crop from all other years and sowing times did not reach a value of 5.00 MJ NE l kg-1, but approach a level of 4.4 to 4.5 MJ NE l kg-1, thus they would not be suitable for preserved feed. It can be recommended that this late season crop should rather be used for grazing of livestock than as preserved feed.

Mastitis is a highly prevalent disease, which negatively affects cow performance, profitability, welfare, and longevity. The objectives of this study were (1) to quantify the impact of the first instance of mastitis, at different stages of lactation, on production and economic performance, and (2) to further quantify the impact of the first instance of mastitis when only cows that remain in the herd for at least 100 d in milk (DIM) and those that remain for 305 DIM are included in the analysis. A retrospective longitudinal study was conducted using data from existing animal health record files and Dairy Herd Improvement records. After editing based on selected inclusion criteria and completeness of health records, data consisted of records from first-lactation Holstein cows, from 120 herds, that calved for the first time between 2003 and 2014, inclusive. Mastitic cows were assigned to 1 of 4 groups based on when in the lactation the first event of mastitis occurred: transition (1–21 DIM), early lactation (22–100 DIM), mid lactation (101–200 DIM), or late lactation (201+ DIM). Mid-lactation and late-lactation mastitic cows were also stratified by cumulative milk yield before the mastitis event. Healthy cows (i.e., no recorded mastitis event) were randomly assigned for each lactation stage, with mid-lactation healthy and late-lactation healthy cows similarly stratified. Production performance (cumulative milk, fat, and protein yield) and economic performance [milk value, margin over feed cost (MOFC), and gross profit] were analyzed using a mixed model with herd as a random effect. Significant losses in cumulative milk yield (−382 to −989 kg) and correspondingly lower fat and protein yields were found in mastitic cows, with transition and late-lactation mastitic cows having the highest losses. Drops in production translated to significant reductions in cumulative milk value (−Can$287 to −Can$591; −US$228 to −US$470), MOFC (−Can$243 to −Can$540; −US$193 to −US$429), and gross profit (−Can$649 to −Can$908; −US$516 to −US$722) for mastitic cows at all stages. Differences between mastitic and healthy cows in the early lactation and transition stages remained for all variables in the 100-DIM analysis, but, aside from gross profit, were nonsignificant in the 305-DIM analysis. Gross profit accounted for all costs associated with mastitis and thus continued to be lower for mastitic cows at all stages, even in the 305-DIM analysis in which culled cows were omitted (−Can$485 to −Can$979; −US$386 to −US$779). The research reflects the performance implications of mastitis, providing more information upon which the producer can make informed culling decisions and maximize both herd profitability and cow longevity.

Precision livestock farming (PLF) involves the use of sensors that captures large amounts of real-time information at the building, herd or animal level, which are later processed to control the system. Data processing can be accomplished using mathematical models (MM), artificial intelligence algorithms (AI) or a combination of these and other methods. The choice of the method must be made according to the volume of data to be processed, its nature and the relationship between the available information and the desired control of the system. Several components of PLF such as precision nutrition, early disease detection, animal welfare among others may require sophisticated data processing methods. MM is today the preferred method to estimate nutrient requirements in the precision nutrition component of PLF. Conventional MM estimate average population responses using historical population information. Important limitations of these models are the assumption that all the individuals of the population have the same response to a given nutrient provision and that they have not been developed for real-time estimations using up-to-date available information. Therefore, MM have to be developed specifically for PLF and operate in real-time at individual or small group level, considering the between and within-animal variation. Growth patterns, nutrient utilization and behavior vary among animals and herds. There are opportunities to combine data-driven AI with knowledge-driven MM to control more complex PLF components. AI thrive in large complex datasets, where establishing connections can be otherwise difficult due to data complexity, volume and where flexibility is needed to process real-time data from individuals. In contrast, knowledge-driven MM can simplify complex biological systems based on well-established concepts and information. In both cases, PLF models must be flexible enough to consider changes over time for the same animal or herd, and among animals and herds, acknowledging the method limitation while using its strength.

The use of fatty acid profiles from milk recording samples to predict body weight change of dairy cows in early lactation in commercial dairy farms Franziska Dettmann12, Daniel Warner1, Albert Johannes Buitenhuis3, Morten Kargo34, Anne Mette Hostrup Kjeldsen4, Niels Henning Nielsen5, Daniel M. Lefebvre1, Debora E. Santschi1 1Valacta, Sainte-Anne-de-Bellevue, Canada 2LKV Niedersachsen e.V., Leer, Germany 3Molecular Biology and Genetics, Aarhus University, Tjele, Denmark 4SEGES, Aarhus, Denmark 5RYK, Aarhus, Denmark

The relationship between in vitro rumen CH4 production of grass silages, using the gas production technique, and in vivo data obtained with the same cows and rations in respiration chambers was investigated. Silages were made from grass harvested in 2013 on May 6th, May 25th, July 1st and July 8th. The grass silages were used to formulate four different rations which were fed to 24 cows in early and late lactation, resulting in a slightly different dry matter intake (DMI; 16.5 kg/day vs. 15.4 kg/day). The experimental rations consisted of 70% grass silage, 10% maize silage, and 20% concentrates on a dry matter basis. Cows were adapted to the rations for 17 days before rumen fluid was collected via oesophageal tubing, and in vitro gas and CH4 production were analysed. In vitro total gas and CH4 production of the (ensiled) grass expressed as ml/g OM decreased with advancing maturity of the grass. The in vitro CH4 production after 48 hr of incubation expressed in ml/g OM did not correlate with the in vivo CH4 production expressed in g/kg organic matter intake or g/kg DMI (R2  = .00-.18, p ≥ .287). The differences in CH4 emission per unit of intake observed in vivo were rather small between the different rations, which also contributed to the observed poor relationship. Utilizing stepwise multiple regression improved the correlation only slightly. In vitro gas and CH4 production varied based on whether donor cows were previously adapted to the respective ration or not, suggesting that careful adaption to the experimental diet should be envisaged in in vitro gas and CH4 production experiments.

The potential of an in vitro gas production (GP) system to predict the in vivo enteric methane (CH4) production for various ryegrass-based silages was evaluated, using adapted rumen fluid from cows. Rumen fluid from 12 lactating rumen-cannulated Holstein-Friesian cows were used for in vitro incubations and compared with in vivo CH4 production data derived from the same cows fed the same grass silages. The cows consumed a total mixed ration consisting of six different grass silages and concentrate at an 80:20 ratio on a dry matter (DM) basis. The grass silages differed in plant maturity at harvest (28, 41 and 62days of regrowth) and N fertilisation (65 and 150kg of N/ha). Rumen fluid from cows consuming each of the six grass silages was used to determine the in vitro organic matter (OM) fermentation and in vitro CH4 synthesis, using an automated GP technique. In vitro GP decreased with increasing maturity of the grass. In vitro CH4 production, expressed either in ml/g of OM, in ml/g of degraded OM (DOM) or as a% of the total GP, increased with increased N fertilisation (P<0.05). Maturity of grass at harvest did not affect the CH4 synthesis expressed in ml/g of DOM and CH4 expressed as% of the total gas, whereas N fertilisation increased the in vitro CH4 synthesis, expressed in any unit. The in vitro data correlated poorly with the in vivo data. Across the six grass silages tested, the in vitro CH4 production, expressed in ml/g of OM after 8, 12, 24, and 72h of incubation did not correlate with the in vivo enteric CH4 production, expressed in g/kg of DM intake (R2=0.01–0.08). Stepwise multiple regression showed a weak, but positive correlation between the observed in vivo CH4 synthesis, expressed in g/kg FPCM and the predicted CH4 per kg FPCM, using the amount of in vitro organic matter degraded (R2=0.40; P=0.036). In vitro gas and CH4 parameters did not improve the accuracy of the prediction of the in vivo CH4 data.

The objective of this study was to determine the effect of level of feed intake and quality of ryegrass silage as well as their interaction on enteric methane (CH) emission from dairy cows. In a randomized block design, 56 lactating dairy cows received a diet of grass silage, corn silage, and a compound feed meal (70:10:20 on DM basis). Treatments consisted of 4 grass silage qualities prepared from grass harvested from leafy through late heading stage, and offered to dairy cows at 96 ± 2.4 (mean ± SEM) days in milk (namely, high intake) and 217 ± 2.4 d in milk (namely, low intake). Grass silage CP content varied between 124 and 286 g/kg of DM, and NDF content between 365 and 546 g/kg of DM. After 12 d of adaptation, enteric CH production of cows was measured in open-circuit climate-controlled respiration chambers for 5 d. No interaction between DMI and grass quality on CH emission, or on milk production, diet digestibility, and energy, and N retention was found ( ≥ 0.17). Cows had a greater DMI (16.6 vs. 15.5 kg/d; SEM 0.46) and greater fat- and protein-corrected milk (FPCM) yield (29.9 vs. 25.4 kg/d; SEM 1.24) at high than low intake (both ≤ 0.001). Apparent total-tract nutrient digestibility was not affected ( ≥ 0.08) by DMI level. Total enteric CH production (346 ± 10.9 g/d) was not affected ( = 0.15) by DMI level. A small, significant ( = 0.025) decrease at high compared with low intake occurred for CH yield (21.8 ± 0.59 g/kg of DMI; -4%). Methane emission intensity (12.8 ± 0.56 g/kg of FPCM; -12%) was considerably smaller ( ≤ 0.001) at high intake as a result of greater milk yields realized in early lactation. As grass quality decreased from leafy through late heading stage, FPCM yield and apparent total-tract OM digestibility declined (-12%; ≤ 0.015), whereas total CH production (+13%), CH yield (+21%), and CH emission intensity (+28%) increased ( ≤ 0.001). Our results suggest that improving grass silage quality by cutting grass at an earlier stage considerably reduces enteric CH emissions from dairy cows, independent of DMI. In contrast, losses of N in manure increased for the earlier cut grass silage treatments. The small increase in DMI at high intake was associated with a small to moderate reduction in CH emission per unit of DMI and GE intake. This study confirmed that enteric CH emissions from dairy cows at distinct levels of feed intake depend on the nutritive value and chemical composition of the grass silage.

The objective of this study was to examine the effects of supplementing increasing amounts of linseed oil (LO) on intake, milk production, and enteric CH4 emissions of dairy cows fed corn silage–based diets. Twelve lactating, multiparous Holstein cows (84 ± 28 d in milk and 42 ± 4.6 kg/d milk yield) were used in a replicated 4 × 4 Latin square design (35-d periods and 14 d of adaptation). Cows were fed ad libitum (5% orts, on an as-fed basis) a corn silage–based TMR (61:39 forage:concentrate ratio) not supplemented (control) or supplemented with 2, 3, or 4% LO (on a DM basis). Methane production was determined (3 consecutive days) using respiration chambers, and intake and milk yield were measured over 6 consecutive days. Data were analyzed using the MIXED procedure (SAS) and differences among treatments were declared significant at P ≤ 0.05 using Dunnett's comparison test. Dry matter intake and energy-corrected milk (ECM) were not affected (23.5 and 33.1 kg/d, respectively) by supplementing LO at 2 and 3%, but they decreased (21.1 and 30.4 kg/d, respectively) when LO was added at 4%. Daily CH4 emission averaged 515 g/d for cows fed the control diet and decreased by 8, 21, and 33% in cows fed 2, 3, and 4% LO, respectively. When adjusted for DMI, CH4 emission averaged 21.7 g/kg for cows fed the control diet and declined in cows fed 2, 3, and 4% LO (19.7, 17.4, and 15.7 g/kg, respectively). When expressed per kilogram of ECM, CH4 production was not affected by supplementing 2% LO (15.2 g/kg) but declined when LO was added at 3 and 4% (12.6 and 11.5 g/kg, respectively). Results of this study show that supplementing a corn silage–based diet with up to 3% of LO reduces enteric CH4 production without adverse effects on DMI and milk production. However, a higher supplementation level (4%) impairs DMI and milk yield. These findings suggest that LO supplementation level should not exceed 3% (DM basis) in corn silage–based diets to mitigate enteric CH4 without negatively affecting animal production.