Search Results (306)

Stereotypic behaviors are common in farm animals and often signal poor welfare. Tongue-rolling is the most prevalent stereotypic behavior in cows. In this study, we compared the plasma and milk composition of 16 high-frequency tongue-rolling cows (HTR group) and 16 non-stereotypic cows (CON group). All cows were primiparous cows. Biochemical tests, plasma metabolomics, and milk proteomics revealed higher plasma triiodothyronine levels in HTR cows, and lower levels of αs1-casein, β-casein, κ-casein, and lactoferrin in their milk. Multi-omics analyses identified 103 differential metabolites and 73 differential proteins, including various GTP-binding proteins, with the Ras signaling pathway being significantly upregulated in the HTR cows. GO enrichment analysis highlighted significant changes in molecular function, particularly related to GTP/GDP-binding proteins. Additionally, HTR cows exhibited elevated cellular metabolic activity. These findings suggest that high-frequency tongue-rolling is associated with altered endocrine and metabolic profiles, disrupted milk protein synthesis, and impaired immune function potential. The reduction in key milk proteins and lactoferrin may negatively impact milk quality and immune defense. Further research is needed to clarify the causal relationship between these physiological changes and tongue-rolling, providing insights into the underlying mechanisms of stereotypic behaviors in dairy cows and their implications for animal welfare and milk production. Full article

This study evaluated six supervised machine learning (ML) models for early detection of subclinical ketosis and low-grade ruminal acidosis in dairy cows during the transition period. Ninety-four Holstein cows were monitored for 21 days postpartum using in-line milk analyzers and intraruminal sensors that continuously recorded milk composition, behavioral, and physiological parameters. Based on clinical examination, blood β-hydroxybutyrate concentration, and fat-to-protein ratio, cows were classified as healthy (n = 44), subclinical ketosis (n = 24), or subclinical acidosis (n = 26). Among the tested models, Random Forest and XGBoost achieved perfect accuracy within this dataset, while Logistic Regression reached 89.5%, Decision Tree 84.2%, and both Naive Bayes and Support Vector Machine 78.9%. These results suggest that ensemble approaches, particularly Random Forest and XGBoost, show strong potential for integration with precision livestock technologies, but their apparent performance should be interpreted cautiously and confirmed in larger, multi-farm studies. Full article

Cow behavior recognition constitutes a fundamental element of effective cow health monitoring and intelligent farming systems. Within large-scale cow farming environments, several critical challenges persist, including the difficulty in accurately capturing behavioral feature information, substantial variations in multi-scale features, and high inter-class similarity among different cow behaviors. To address these limitations, this study introduces an enhanced target detection algorithm for cow behavior recognition, termed RFR-YOLO, which is developed upon the YOLOv11n framework. A well-structured dataset encompassing nine distinct cow behaviors—namely, lying, standing, walking, eating, drinking, licking, grooming, estrus, and limping—is constructed, comprising a total of 13,224 labeled samples. The proposed algorithm incorporates three major technical improvements: First, an Inverted Dilated Convolution module (Region Semantic Inverted Convolution, RsiConv) is designed and seamlessly integrated with the C3K2 module to form the C3K2_Rsi module, which effectively reduces computational overhead while enhancing feature representation. Second, a Four-branch Multi-scale Dilated Attention mechanism (Four Multi-Scale Dilated Attention, FMSDA) is incorporated into the network architecture, enabling the scale-specific features to align with the corresponding receptive fields, thereby improving the model’s capacity to capture multi-scale characteristics. Third, a Reparameterized Generalized Residual Feature Pyramid Network (Reparameterized Generalized Residual-FPN, RepGRFPN) is introduced as the Neck component, allowing for the features to propagate through differentiated pathways and enabling flexible control over multi-scale feature expression, thereby facilitating efficient feature fusion and mitigating the impact of behavioral similarity. The experimental results demonstrate that RFR-YOLO achieves precision, recall, mAP50, and mAP50:95 values of 95.9%, 91.2%, 94.9%, and 85.2%, respectively, representing performance gains of 5.5%, 5%, 5.6%, and 3.5% over the baseline model. Despite a marginal increase in computational complexity of 1.4G, the algorithm retains a high detection speed of 147.6 frames per second. The proposed RFR-YOLO algorithm significantly improves the accuracy and robustness of target detection in group cow farming scenarios. Full article

In real-world dairy farming environments, object recognition models often suffer from missed or false detections due to complex backgrounds and cow occlusions. In response to these issues, this paper proposes FSCA-YOLO, a multi-object cow behavior recognition model based on an improved YOLOv11 framework. First, the FEM-SCAM module is introduced along with the CoordAtt mechanism to enable the model to better focus on effective behavioral features of cows while suppressing irrelevant background information. Second, a small object detection head is added to enhance the model’s ability to recognize cow behaviors occurring at the distant regions of the camera’s field of view. Finally, the original loss function is replaced with the SIoU loss function to improve recognition accuracy and accelerate model convergence. Experimental results show that compared with mainstream object detection models, the improved YOLOv11 in this section demonstrates superior performance in terms of precision, recall, and mean average precision (mAP), achieving 95.7% precision, 92.1% recall, and 94.5% mAP—an improvement of 1.6%, 1.8%, and 2.1%, respectively, over the baseline YOLOv11 model. FSCA-YOLO can accurately extract cow features in real farming environments, providing a reliable vision-based solution for cow behavior recognition. To support specific behavior recognition and in-region counting needs in multi-object cow behavior recognition and tracking systems, OpenCV is integrated with the recognition model, enabling users to meet the diverse behavior identification requirements in groups of cows and improving the model’s adaptability and practical utility. Full article

The feeding behavior of dairy cows constitutes a complex temporal sequence comprising actions such as head lowering, sniffing, arching, eating, head raising, and chewing. Its precise recognition is crucial for refined livestock management. While existing 2D convolution-based models effectively extract features from individual frames, they lack temporal modeling capabilities. Conversely, due to their high computational complexity, 3D convolutional networks suffer from significantly limited recognition accuracy in high-density feeding scenarios. To address this, this paper proposes a Spatio-Temporal Fusion Network (DAS-Net): it designs a collaborative architecture featuring a 2D branch with a triple-attention module to enhance spatial key feature extraction, constructs a 3D branch based on multi-branch dilated convolution and integrates a 3D multi-scale attention mechanism to achieve efficient long-term temporal modeling. On our Spatio-Temporal Dairy Feeding Dataset (STDF Dataset), which contains 403 video clips and 10,478 annotated frames across seven behavior categories, the model achieves an average recognition accuracy of 56.83% for all action types. This result marks a significant improvement of 3.61 percentage points over the original model. Among them, the recognition accuracy of the eating action has been increased to 94.78%. This method provides a new idea for recognizing dairy cow feeding behavior and can provide technical support for developing intelligent feeding systems in real dairy farms. Full article

For monitoring animal adaptation when facing environmental challenges, and more specifically when addressing the impacts of global warming—particularly responses to heat stress and short-term fluctuations in osmotic regulations in the different organs influencing animal physiology—there is an increasing demand for digital tools to understand and monitor a range of biomarkers. Microneedle arrays (MNAs) have recently emerged as promising devices minimally invasively penetrating human skin to access dermal interstitial fluid (ISF) to monitor deviations in physiology and consequences on health. The ISF is a blood filtrate where the concentrations of ions, low molecular weight metabolites (<70 kDa), hormones, and drugs, often closely correlate with those in blood. However, anatomical skin differences between human and farm animals, especially large animals, as well as divergent tolerances of such devices among species with behavior specificities, motivate new MNA designs. We addressed technological challenges to design higher microneedles for farm animal (pigs and cattle) measurements. We designed microneedle arrays composed of 37 microneedles, each 2.8 mm in height, using dextran-methacrylate, a photo-crosslinked biocompatible biopolymer-based hydrogel. The arrays were characterized geometrically and mechanically. Their abilities to perforate pig and cow skin were demonstrated through histological analysis. The MNAs successfully absorbed approximately 10 µL of fluid within 3 h of application. Full article

Individual cow identification is a prerequisite for automatically monitoring behavior patterns, health status, and growth data of each cow, and can provide the assistance in selecting excellent cow individuals for breeding. Despite high recognition accuracy, traditional implantable electronic devices such as RFID (i.e., Radio Frequency Identification) can cause some degree of harm or stress reactions to cows. Image-based methods are widely used due to their non-invasive advantages, but these methods have poor adaptability to different environments and target size, and low detection accuracy in complex scenes. To solve these issues, this study designs a Dy_Conv (i.e., dynamic convolution) module and innovatively constructs a Dynamic_Bottleneck module based on the Dy_Conv and S2Attention (Sparse-shift Attention) mechanism. On this basis, we replaces the first and fourth bottleneck layers of Resnet50 with the Dynamic_Bottleneck to achieve accurate extraction of local features and global information of cows. Furthermore, the QAConv (i.e., query adaptive convolution) module is introduced into the front end of the backbone network, and can adjust the parameters and sizes of convolution kernels to adapt to the scale changes in cow targets and input images. At the same time, NAM (i.e., normalization-based attention module) attention is embedded into the backend of the network to achieve the feature fusion in the channels and spatial dimensions, which contributes to better distinguish visually similar individual cows. The experiments are conducted on the public datasets collected from different cowsheds. The experimental results showed that the Rank-1, Rank-5, and mAP metrics reached 96.8%, 98.9%, and 95.3%, respectively. Therefore, the proposed model can effectively capture and integrate multi-scale features of cow body appearance, enhancing the accuracy of individual cow identification in complex scenes. Full article

This study aimed to evaluate the accuracy of the accelerometer-equipped collar RUMI to detect estrus in dairy cows, establish a recommendation for the optimal timing for artificial insemination (AI) when using this device, and characterize the blood flow of the dominant follicle (F) and the corpus luteum (CL) as ovulation approaches. Forty-seven cycling cows were monitored following synchronization with a modified G6G protocol, allowing for spontaneous ovulation. Ultrasound examinations were conducted every 12 h, starting 48 h after the second PGF2α dose, to monitor uterine and ovarian changes. Blood samples were also collected to determine serum progesterone (P4) levels. Each cow was fitted with a RUMI collar, which continuously monitored behavioral changes to identify the onset, offset, and peak of activity of estrus. One-way ANOVA assessed the relationship between physiological parameters and time before ovulation. Results showed that the RUMI collar demonstrated high specificity (100%), sensitivity (90.90%), and accuracy (93.62%) for estrus detection. The optimal AI window was identified as between 11.4 and 15.5 h after heat onset. Increased blood flow to the F and reduced luteal activity were observed in the 48 h prior to ovulation. Further research is needed to assess the influence of this AI window on conception rates, and if it should be modified considering external factors. Full article

The ‘head back’ posture is a pronounced and significant behavioral trait during bovine parturition, commonly interpreted as a natural response to the pain associated with parturition. Leveraging computer vision technology for real-time monitoring of parturition behaviors can provide timely assistance during calving and enhance animal welfare. This study initially evaluated the head back posture in cows of different types, finding that primiparous cows and those delivering calves weighing over 43 kg exhibited prolonged durations of both labor and head back posture. A model was developed using the YOLOv8 algorithm with 25,617 images to recognize and classify changes in head posture during parturition, including positions like lying with or without head back. The model demonstrated robust predictive performance with a precision (P) of 69.76%, recall (R) of 75.35%, average precision (AP) of 70.12%, and F₁ score of 0.71. Furthermore, the model’s capability to recognize postures from different camera angles and under varying environmental conditions was assessed. Notably, images captured from an abdominal angle achieved AP exceeding 90%, with consistent stability under varying lighting conditions, including sunny and overcast weather, during both daytime and nighttime. Behavioral analysis showed that the parturition duration and total duration of head back posture in primiparous cows were significantly higher than those in multiparous cows (p < 0.05), and the changing trends of motor performance between primiparous and multiparous cows were consistent across different parturition stages. Additionally, the correlation between calf birth weight and maternal behavior was stronger in primiparous cows than in multiparous cows, further indicating obvious differences in physiological and behavioral responses of cows during primiparous and multiparous parturition. This study underscores the potential of computer vision applications in enhancing real-time intervention and promoting welfare during bovine parturition. Full article

Heat stress is one of the main welfare and productivity problems faced by dairy cattle in Mediterranean climates. The main objective of this work is to predict heat stress in livestock from shade-seeking behavior captured by computer vision, combined with some climatic features, in a completely non-invasive way. To this end, we evaluate two soft computing algorithms—Random Forests and Neural Networks—clarifying the trade-off between accuracy and interpretability for real-world farm deployment. Data were gathered at a commercial dairy farm in Titaguas (Valencia, Spain) using overhead cameras that counted cows in the shade every 5–10 min during summer 2023. Each record contains the shaded-cow count, ambient temperature, relative humidity, and an exact timestamp. From here, three thermal indices were derived: the current THI, the previous-night mean THI, and the day-time accumulated THI. The resulting dataset covers 75 days and 6907 day-time observations. To evaluate the models’ performance a 5-fold cross-validation is also used. The results show that both soft computing models outperform a single Decision Tree baseline. The best Neural Network (3 hidden layers, 16 neurons each, learning rate $={10}^{-3}$) reaches an average RMSE of $14.78$, while a Random Forest (10 trees, depth $=5$) achieves $14.97$ and offers the best interpretability. Daily error distributions reveal a median RMSE of $13.84$ and confirm that predictions deviate less than one hour from observed shade-seeking peaks. Although the dataset came from a single farm, the results generalized well within the observed range. However, the models could not accurately predict the exact number of cows in the shade. This suggests the influence of other variables not included in the analysis (such as solar radiation or wind data), which opens the door for future research. Full article

This study evaluated the effects of reduced nitrogen fertilization and the intercropping of annual ryegrass (Lolium multiflorum Lam.) with forage legumes—common vetch (Vicia sativa L.) and red clover (Trifolium pratense L.)—on milk production and enteric methane emissions in grazing dairy cows. Twelve Holstein × Jersey cows were assigned to a crossover design involving two treatments: ryegrass monoculture (RG) or ryegrass—legume mixture (RG + Leg). Methane emissions were measured using GreenFeed systems; grazing behavior, milk yield and composition, and organic matter digestibility were also assessed. Legume inclusion contributed ~9% of the pre-grazing biomass, and cows grazing RG + Leg pastures had lower herbage mass (−214 kg DM/ha) and lower herbage allowance (−6 kg DM/cow/day) than cows on monoculture ryegrass. Energy-corrected milk (ECM), methane emissions (g/day and g/kg ECM), and grazing behavior were not significantly affected by treatment. These results suggest that, under subtropical grazing conditions, reducing nitrogen fertilization combined with the modest inclusion of vetch and red clover does not mitigate enteric methane emissions nor enhance animal performance. Enhanced strategies to increase legume proportion in mixed swards are needed to unlock their potential for sustainable intensification of pasture-based dairy systems. Full article

Modern beef cattle reproductive management faces increasing challenges due to the growing global demand for beef. Reproductive efficiency is a critical factor determining the productivity and profitability of beef cattle operations. Optimal reproductive performance in a beef cattle herd is achieved when each cow produces one calf per year, maintaining a calving interval of 365 days. However, this goal is difficult to achieve, as the gestation period in beef cows lasts approximately 280 days, leaving only 80–85 days for successful conception. Traditional methods, such as visual estrus detection, are becoming increasingly unreliable due to expanding herd sizes and the subjectivity of visual observation. Additionally, silent estrus—where ovulation occurs without noticeable behavioral changes—further complicates the accurate estrous-based identification of the optimal insemination period. To enhance reproductive efficiency, advanced technologies are increasingly being integrated into cattle management. Sensor-based monitoring systems, including accelerometers, pedometers, and ruminoreticular boluses, enable the precise tracking of activity changes associated with the estrous cycle. Furthermore, infrared thermography offers a non-invasive method for detecting body temperature fluctuations, allowing for more accurate estrus identification and optimized timing of insemination. The use of these innovative technologies has the potential to significantly improve reproductive efficiency in beef cattle herds and contribute to overall farm productivity and sustainability. The objective of this review is to examine advancements in smart technologies applied to beef cattle reproductive management, presenting commercially available technologies and recent scientific studies on innovative systems. The focus is on sensor-based monitoring systems and infrared thermography for optimizing reproduction. Additionally, the challenges associated with these technologies and their potential to enhance reproductive efficiency and sustainability in the beef cattle industry are discussed. Despite the benefits of advanced technologies, their implementation in cattle farms is hindered by financial and technical challenges. High initial investment costs and the complexity of data analysis may limit their adoption, particularly in small and medium-sized farms. However, the continuous development of these technologies and their adaptation to farmers’ needs may significantly contribute to more efficient and sustainable reproductive management in beef cattle production. Full article

This study aimed to evaluate daily patterns of hourly milking frequency (MF) in dairy cows milked with an automatic milking system (AMSs), considering the effects of season, parity order (PO), days in milk (DIM), and milk yield (MY). A retrospective longitudinal study was conducted on a commercial dairy farm in southern Brazil over one year using data from 130 Holstein cows and 94,611 milking events. MF data were analyzed using general linear models. Overall, hourly MF followed a consistent daily pattern, with peaks between 4:00 and 11:00 a.m. and between 2:00 and 6:00 p.m., regardless of season, PO, DIM, or MY category. MF was higher in primiparous (2.84/day, p = 0.0013), early-lactation (<106 DIM; 3.00/day, p < 0.0001), and high-yielding cows (≥45 L/day; 3.09/day, p < 0.0001). High-yielding cows also showed sustained milking activity into the late nighttime. Although seasonal and individual factors significantly affected MF, they had limited influence on the overall daily distribution of milkings. These results suggest stable behavioral patterns within the specific AMS management conditions observed in this study and suggest that adjusting milking permissions and feeding strategies based on cow characteristics may improve system efficiency. Full article

Milk lactose concentration has been proposed as a noninvasive indicator of metabolic health in dairy cows, particularly during early lactation when metabolic demands are elevated. This study aimed to investigate the relationship between milk lactose levels and physiological, biochemical, and behavioral parameters in early-lactation Holstein cows. Twenty-eight clinically healthy cows were divided into two groups: Group 1 (milk lactose < 4.70%, n = 14) and Group 2 (milk lactose ≥ 4.70%, n = 14). Both groups were monitored over a 21-day period using the Brolis HerdLine in-line milk analyzer (Brolis Sensor Technology, Vilnius, Lithuania) and SmaXtec intraruminal boluses (SmaXtec Animal Care Technology^®, Graz, Austria). Parameters including milk yield, milk composition (lactose, fat, protein, and fat-to-protein ratio), blood biomarkers, and behavior were recorded. Cows with higher milk lactose concentrations (≥4.70%) produced significantly more milk (+12.76%) and showed increased water intake (+15.44%), as well as elevated levels of urea (+21.63%), alanine aminotransferase (ALT) (+22.96%), glucose (+4.75%), magnesium (+8.25%), and iron (+13.41%) compared to cows with lower lactose concentrations (<4.70%). A moderate positive correlation was found between milk lactose and urea levels (r = 0.429, p < 0.01), and low but significant correlations were observed with other indicators. These findings support the use of milk lactose concentration as a practical biomarker for assessing metabolic and physiological status in dairy cows, and highlight the value of integrating real-time monitoring technologies in precision livestock management. Full article

Virtual fencing (VF) offers a promising alternative to conventional or electrified fences for managing livestock grazing distribution. This study evaluated the behavioral responses of 25 Rarámuri Criollo cows fitted with Nofence^® collars in Pine Valley, CA, USA. The VF system was deployed in chaparral rangeland pastures. The study included a 14-day training phase followed by an 18-day testing phase. The collar-recorded variables, including audio warnings and electric pulses, animal movement, and daily typical behavior patterns of cows classified into a High or Low virtual fence response group, were compared using repeated-measure analyses with mixed models. During training, High-response cows (i.e., resistant responders) received more audio warnings and electric pulses, while Low-response cows (i.e., active responders) had fewer audio warnings and electric pulses, explored smaller areas, and exhibited lower mobility. Despite these differences, both groups showed a time-dependent decrease in the pulse-to-warning ratio, indicating increased reliance on audio cues and reduced need for electrical stimulation to achieve similar containment rates. In the testing phase, both groups maintained high containment with minimal reinforcement. The study found that Rarámuri Criollo cows can effectively adapt to virtual fencing technology, achieving over 99% containment rate while displaying typical diurnal patterns for grazing, resting, or traveling behavior. These findings support the technical feasibility of using virtual fencing in chaparral rangelands and underscore the importance of accounting for individual behavioral variability in behavior-based containment systems. Full article