Milk production in Europe is facing major challenges to ensure its economic, environmental, and social sustainability. It is essential that holistic concepts are developed to ensure the future sustainability of the sector and to assist farmers and stakeholders in making knowledge-based decisions. In this study, integrated sustainability assessment by means of whole-farm modelling is presented as a valuable approach for identifying factors and mechanisms that could be used to improve the three pillars (3Ps) of sustainability in the context of an increasing awareness of economic profitability, social well-being, and environmental impacts of dairy production systems (DPS). This work aims (i) to create an evaluation framework that enables quantitative analysis of the level of integration of 3P sustainability indicators in whole-farm models and (ii) to test this method. Therefore, an evaluation framework consisting of 35 indicators distributed across the 3Ps of sustainability was used to evaluate three whole-farm models. Overall, the models integrated at least 40% of the proposed indicators. Different results were obtained for each sustainability pillar by each evaluated model. Higher scores were obtained for the environmental pillar, followed by the economic and the social pillars. In conclusion, this evaluation framework was found to be an effective tool that allows potential users to choose among whole-farm models depending on their needs. Pathways for further model development that may be used to integrate the 3P sustainability assessment of DPS in a more complete and detailed way were identified.

This study aimed to expand the knowledge about the activity and mode of action of CHI on methanogenesis and rumen microbial populations in vivo. A total of 16 lactating dairy cows were distributed in two groups, one of them receiving 135 mg CHI/kg body weight daily. The effect on productive performance, milk composition, fermentation efficiency, methane emissions, microbial protein synthesis, and ruminal microbial communities was determined. Supplementation with CHI did not affect rumen microbial diversity but increased the relative abundance (RA) of the bacteria Anaeroplasma and decreased those of rumen ciliates and protozoa resulting in a shift towards a lower acetic to propionic ratio. However, no effect on milk yield or methane intensity was observed. In conclusion, supplementing 135 mg CHI/kg body weight increased the RA of Anaeroplasma and decreased those of rumen ciliates and protozoa, both being related to fiber degradation in the rumen in different ways and resulted in a shift of ruminal fermentation towards more propionate proportions, without affecting CH₄ emissions, milk yield, or milk composition. Further research with higher doses would be necessary to assess the potential use of this additive as a methane inhibitor.

European dairy production faces significant economic, environmental, and social sustainability challenges. Given the great diversity of dairy cattle production systems in Europe, region-specific concepts to improve environmental and socioeconomic sustainability are needed. Regionally integrated dairy cattle-crop systems emerge as a more resilient and sustainable alternative to highly specialized farming systems. Identifying different dairy cattle production typologies and their potential interactions with fodder crop production is presented as a step in transitioning to optimized agricultural systems. Currently existing typologies of integrated systems are often insufficient when characterizing structural, socioeconomic, and environmental components of farms. We fill this gap in the literature by identifying, describing, and comparing representative dairy cattle production system typologies and their interrelation with regional fodder crop production at the European regional scale. This is a necessary step to assess the scope for adapted mitigation and sustainability measures in the future. For this purpose, a multivariate statistical approach is applied. We show how different land-use practices, farm structure characteristics, socio-economic attributes, and emission intensities condition dairy production. Furthermore, the diversity of regional fodder crop production systems is demonstrated by analyzing their distribution in Europe. Together with identified typologies, varying degrees of regional specialization in milk production allow for identifying future strategies associated with the application of integrated systems in key European dairy regions. This study contributes to a better understanding of the existing milk production diversity in Europe and their relationship with regional fodder crop production. In addition, we discuss the benefits of integrated systems as a clear, viable, and resilient alternative to ongoing livestock intensification in the European context. Identifying interactions between components of integrated systems will facilitate decision-making, the design and implementation of measures to mitigate climate change, and the promotion of positive socio-economic and environmental interactions.

Understanding the environmental consequences associated with dairy cattle production systems is crucial for the implementation of targeted strategies for emission reduction. However, few studies have modelled the effect of tailored emission mitigation options across key European dairy production systems. Here, we assess the single and combined effect of six emission mitigation practises on selected case studies across Europe through the Sustainable and Integrated Management System for Dairy Production model. This semi-mechanistic model accounts for the interacting flows from a whole-farm perspective simulating the environmental losses in response to different management strategies and site-specific conditions. The results show how reducing the crude protein content of the purchased fraction of the diet was an adequate strategy to reduce the greenhouse gas and nitrogen emission intensity in all systems. Furthermore, implementing an anaerobic digestion plant reduced the greenhouse gas emissions in all tested case studies while increasing the nitrogen emissions intensity, particularly when slurry was applied using broadcast. Regarding the productivity increase, contrasting effects were observed amongst the case studies modelled. Moreover, shallow slurry injection effectively mitigated the intensity of nitrogen losses from the fields due to strong reductions in ammonia volatilisation. When substituting urea with ammonium nitrate as mineral fertiliser, site-specific conditions affected the mitigation potential observed, discouraging its application on sandy-loam soils. Rigid slurry covers effectively reduced the storage-related nitrogen emissions intensity while showing a minor effect on total greenhouse gas emission intensity. In addition, our results provide novel evidence regarding the advantages of cumulative implementation of adapted mitigation options to offset the negative trade-offs of single-option applications (i.e. slurry covers or anaerobic digestion and slurry injection). Through this study, we contribute to a better understanding of the effect of emission mitigation options across dairy production systems in Europe, thus facilitating the adoption of tailored and context-specific emission reduction strategies.

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.