Understanding intake on pastures: how, why, and a way forward

Machine Learning Methods and Visual Observations to Categorize Behavior of Grazing Cattle Using Accelerometer Signals

Accelerometers worn by animals produce distinct behavioral signatures, which can be classified accurately using machine learning methods such as random forest decision trees. The objective of this study was to identify accelerometer signal separation among parsimonious behaviors. We achieved this objective by (1) describing functional differences in accelerometer signals among discrete behaviors, (2) identifying the optimal window size for signal pre-processing, and (3) demonstrating the number of observations required to achieve the desired level of model accuracy,. Crossbred steers (Bos taurus indicus; n = 10) were fitted with GPS collars containing a video camera and tri-axial accelerometers (read-rate = 40 Hz). Distinct behaviors from accelerometer signals, particularly for grazing, were apparent because of the head-down posture. Increasing the smoothing window size to 10 s improved classification accuracy (p < 0.05), but reducing the number of observations below 50% resulted in a decrease in accuracy for all behaviors (p < 0.05). In-pasture observation increased accuracy and precision (0.05 and 0.08 percent, respectively) compared with animal-borne collar video observations.

On-farm evaluation of models to predict herbage intake of dairy cows grazing temperate semi-natural grasslands

The objective of the present on-farm study was to evaluate the adequacy of existing models in predicting the pasture herbage DM intake (PDMI) of lactating dairy cows grazing semi-natural grasslands. The prediction adequacy of 13 empirical and semi-mechanistic models, which were predominantly developed to represent stall-fed cows or cows grazing high-quality pastures, were evaluated using the mean bias, relative prediction error (RPE), and partitioning of mean square error of prediction, where models with an RPE ≤ 20% were considered adequate. The reference dataset comprised n = 233 individual animal observations from nine commercial farms in South Germany with a mean milk production, DM intake, and PDMI (arithmetic means ± one SD) of 24 kg/d, (±5.6), 21 kg/d (±3.2), and 12 kg/d (±5.1), respectively. Despite their adaptation to grazing conditions, the behaviour-based and semi-mechanistic grazing-based models had the lowest prediction adequacy among the evaluated models. Their underlying empirical equations likely did not fit the grazing and production conditions of low-input farms using semi-natural grasslands for grazing. The semi-mechanistic stall-based model Mertens II with slight modifications achieved the highest and a satisfactory modelling performance (RPE = 13.4%) when evaluated based on the mean observed PDMI, i.e., averaged across animals per farm and period (n = 28). It also allowed for the adequate prediction of PDMI on individual cows (RPE = 18.5%) that were fed < 4.8 kg DM of supplement feed per day. Nevertheless, when used to predict PDMI of individual animals receiving a high supplementation level, the model Mertens II also did not meet the threshold for an acceptable adequacy (RPE = 24.7%). It was concluded that this lack of prediction adequacy for animals receiving greater levels of supplementation was due to a lack of modelling precision, which mainly could be related to inter-animal and methodological limitations such as the lack of individually measured supplement feed intake for some cows. The latter limitation is a trade-off of the on-farm research approach of the present study, which was chosen to represent the range in feed intake of dairy cows across the diverse low-input farming systems using semi-natural grasslands for grazing.

Beef Steers and Enteric Methane: Reducing Emissions by Managing Forage Diet Fiber Content

Understanding the methane (CH4) emissions that are produced by enteric fermentation is one of the main problems to be solved for livestock, due to their GHG effects. These emissions are affected by the quantity and quality of their diets, thus, it is key to accurately define the intake and fiber content (NDF) of these forage diets. On the other hand, different emission prediction equations have been developed; however, there are scarce and uncertain results regarding their evaluation of the emissions that have been observed in forage diets. Therefore, the objectives of this study were to evaluate the effect of the NDF content of a forage diet on CH4 enteric emissions, and to evaluate the ability of models to predict the emissions from the animals that are consuming these forage diets. In total, thirty-six Angus steers (x¯ = 437 kg live weight) aged 18 months, blocked by live weight and placed in three automated feeding pens, were used to measure the enteric CH4. The animals were randomly assigned to two forage diets (n = 18), with moderate (<50%, MF) and high (>50%, HF) NDF contents. Their dry matter intake was recorded individually, and the CH4 emissions were measured using the SF6 tracer gas technique. For the model evaluation, six prediction equations were compared with 29 studies (n = 97 observations), analyzing the accuracy and precision of their estimates. The emission intensities per unit of DMI, per ADG, and per gross energy intake were significantly lower (p < 0.05) in the animals consuming the MF diet than in the animals consuming the HF diet (21.7 vs. 23.7 g CH4/kg DMI, 342 vs. 660 g CH4/kg ADG, and 6.7% vs. 7.5%, respectively), but there were no differences in the absolute emissions (p > 0.05). The best performing model was the IPCC 2006 model (r2 = 0.7, RMSE = 74.04). These results show that reducing the NDF content of a forage diet by at least 10% (52 g/kg DM) reduces the intensity of the g CH4/kg DMI by up to 8%, and that of the g CH4/kg ADG by almost half. The use of the IPCC 2006 model is suitable for estimating the CH4 emissions from animals consuming forage-based diets.

Grassland-based ruminant farming systems in China: Potential, challenges and a way forward

With an increasing demand for high-quality, eco-friendly food products and growing concerns over ecological conservation, the development of ecology-based alternatives for ruminant production in China is urgently needed. This review discusses the capabilities for integrating grassland grazing into existing livestock farming systems to meet the contemporary human needs for high-quality foods and ecologically stable environments. Additionally, this review provides a critical analysis of the challenges and future directions associated with grassland-based ruminant farming systems. Integrating nutritional manipulation with grazing manipulation is critical for improving the productivity of grassland-based ecosystems and natural ecological functions. Biodiversity is the primary determinant of grassland ecosystem functions, while the composition and function of rumen microbiomes determine ruminant production performance. Future studies should focus on the following aspects: 1) how livestock grazing regulates grassland biodiversity and the mechanisms of grassland biodiversity maintenance, offering an important scientific basis for guiding grazing manipulation practices, including grazing intensity, livestock types, and grazing management practices; to 2) characterize the microbial ecology within the rumen of grazing ruminants to offer clarified instruction for the nutritional manipulation of grazing ruminants. Our recommendation includes creating a transdisciplinary system that integrates ecology, animal nutrition, and animal behavior to develop grassland-based ruminant farming systems sustainably, thereby achieving high-quality animal production and environmentally sustainable goals.

Characterization of grazing behaviour microstructure using point-of-view cameras

Grazing patterns, intake structure, and diet selection are dynamic responses to animals' feeding environment. This study uses video sequences from animal-borne cameras to capture time- and scale-dependent grazing behaviour variables related to sward explanatory conditions. We observed grazing 'through' the sheep's eyes using point-of-view (POV) cameras coupled with event logging software. Time-specific sward features were measured by sampling 'really' grazed patches identified by applying a global navigation satellite system (GNSS) precision-grazing approach. Sward variables on a Mediterranean native sward were measured for two years during the active spring plant-growth cycle. Overall, the results demonstrate that POV cameras were able to capture grazing behaviour fine-tuning to changes in sward characteristics. Sheep compensate for the decrease in sward quantity and nutritive value by increasing the size and duration at each behavioural scale (i.e., meal, bout, and station) while increasing the bout rate and decreasing the station rate. Diet composition also changed as sward matured. The proportion of forbs in the diet remained high in early and late spring, and forbs and legumes were preferred to grasses in early spring. Grazing selectivity was more pronounced in late spring, with sheep favouring the middle stratum of the sward's vertical structure, preferring green vegetative material, while enlarging the feeding niches' span and spending more time at each niche, consequently reducing the station rate. Although data collected by individual animal-borne POV cameras were representative of the flock behaviour, they may underestimate the total grazing time outside major meals. The results indicate that the use of animal-borne video cameras is suitable for assessing variations in sheep grazing behaviour patterns in complex swards.