Improvement of the Welfare Quality scoring model for dairy cows to fit experts' opinion.

After several years of implementation, the original Welfare Quality scoring model for dairy cows appears to be highly sensitive to the number and cleanliness of drinkers and not enough to the prevalence of diseases, and as a consequence may not fit the opinion of some animal welfare experts. The present paper aims to improve the Welfare Quality calculations for the criteria 'Absence of prolonged thirst' and 'Absence of disease' in dairy cows, so that the results are more sensitive to input data and better fit experts' opinion. First, we modified the calculation of 'Absence of prolonged thirst' by linearising the calculation for drinkers' availability to avoid threshold effects. Second, we modified the calculation of 'Absence of disease' by applying a Choquet integral on the three lowest spline-based scores for each health disorder to limit compensation between health disorders. Third, we performed a global sensitivity analysis of the original and the alternative scoring models. Fourth, we compared the results obtained with the original and the alternative models with eight experts' opinions on two subsets composed of 44 and 60 farms, respectively, inspected using the Welfare Quality protocol and on which experts gave their opinion on the overall level of animal welfare. Results show that the alternative model significantly reduced the 'threshold effects' related to the number of drinkers and the compensation between health disorders. On the first subset, the alternative model fits the experts' opinion slightly better than the original model (P = 0.061). On the second subset, the models performed equally. In conclusion, the proposed refinements for calculating scores are validated since they significantly reduced 'threshold effects' and the influence of measures related to drinkers. It also reduced the compensation between health disorders by considering only the three lowest scores and thus increasing the influence of measures related to health disorders, and slightly improve at overall score level the accordance with experts' opinion.

Co-construction of a method for evaluating the intrinsic quality of bovine milk in relation to its fate.

There are time-tested assessments for the environmental and economic aspects of sustainability. Its societal aspect has mainly been approached through the assessment of animal welfare. However, the intrinsic quality of milk is seldom taken into account. We developed a participatory construction method for the overall assessment of intrinsic milk quality in its different dimensions (sensory, technological, nutritional and health), according to the fate of the raw milk. Two assessment models were developed, for semi-skimmed standardized ultra-high temperature (UHT) milk and for pressed uncooked non-standardized raw milk cheese. They were constructed by a participatory approach involving experts in the dairy sector with the aim to obtain a diagnostic tool that could be used in the field to help farmers to manage the quality of their milk (by prioritizing improvements on major problems). They were shaped from prerequisite specifications (limited costs and time of application, desire to obtain a transparent tool with all the steps kept visible) and current technical and scientific knowledge. They were based on indicators obtained from raw bulk tank milk analyses (30 for UHT milk and 50 for cheese assessments), which were then aggregated into criteria, principles, dimensions and overall intrinsic quality at farm level. The assessment models had parts in common, for example, same four dimensions, common indicators for health and nutritional dimensions. They also had process-specific features: units chosen, criteria, indicators and weightings in relation to the final product specifications. For instance, sensory and technological dimensions are more complex and preponderant in the cheese assessment (three principles for cheese vs one for UHT milk in both dimensions). Another example is the lack of microbial pathogens (as potential health risk for consumer) in the UHT milk assessment because of pasteurization. The assessment models then underwent a sensitivity analysis and an application in 30 farms in indoor and grazing periods to finally obtain overall UHT milk and cheese quality scores at a 1-year level. The tool was found to be applicable at farm level. However, we observed low overall quality scores with a narrow dispersion, characteristic of a severe evaluation. Even so, the assessment models showed up seasonal differences of the UHT milk and cheese quality at both overall and dimensional levels. In the light of new scientific knowledge and future quality objectives, these are adaptable to other dairy products allowing for their specific features.

New recommendations for self-locking barriers to reduce skin injuries in dairy cows.

The design of self-locking barriers can affect cows' skin injuries and impair welfare. This study aimed to propose and refine recommendations, expressed relatively to the cows' dimensions, for self-locking barrier design to reduce risks for skin injuries on the neck/shoulder/back and on carpus of dairy cows. We recorded individual body dimensions and the dimensions of self-locking barriers (e.g. top rail height) and assessed skin injuries on 3801 cows from 131 loose-housing dairy farms. We explored the significant associations between presence/absence of skin injuries and self-locking barrier dimensions using weighted multivariable logistic regression, taking into account the diversity of feeding barriers within each farm. The robustness of the models was assessed by cross-validation. Cows had skin injuries mainly on the neck/shoulder/back (29.0%) and, to a lesser extent, on the carpus (14.0%). The final multivariable logistic regression models comprised 13 factors for skin injuries on the neck/shoulder/back, and 11 factors for skin injuries on the carpus. Skin injuries were significantly reduced when the self-locking barriers were inclined (neck/shoulder/back) and when the cows used a feeding table (i.e. flat) instead of a feeding manger or cribs (i.e. hollow) (carpus). A top rail height >1.05 × cow height (measured at withers) was significantly associated with fewer skin injuries on the neck/shoulder/back and on carpus. Skin injuries on the neck/shoulder/back and carpus were significantly reduced when the bottom rail was on the food side relative to the wall, and at a height <0.39 of cow height. Skin injuries were significantly less frequent when the separation wall had no sharp edges on the food side (neck/shoulder/back), was >0.4 of cow height (carpus), was thinner than 15 cm (neck/shoulder/back and carpus) and when the height of the feeding step was 0.04 to 0.1 of cow height (neck/shoulder/back) and the length of the feeding step was <0.2 of cow length (carpus). A headlock articulation nut positioned between 0.62 and 0.78 of cow height significantly reduced skin injuries on the neck/shoulder/back. Here, by combining the diversity of on-farm self-locking barriers and their respective dimensions, we were able to refine the International Commission of Agricultural and Biosystems Engineering recommendations for self-locking barrier design and to propose new ones. This information now needs to be confirmed on other datasets, but can already help farmers and dairy industry stakeholders improve the design of self-locking barriers to improve dairy cow welfare.

Modelling the transfer of 14C from the atmosphere to grass: a case study in a grass field near AREVA-NC La Hague.

Radioactive (14)C is formed as a by-product of nuclear power generation and from operation of nuclear fuel reprocessing plants like AREVA-NC La Hague (North France), which releases about 15 TBq per year of (14)C into the atmosphere. Since the autumn of 2006, (14)C activity concentrations in samples from the terrestrial environment (air, grass and soil) have been monitored monthly on grassland 2 km downwind of the reprocessing plant. The monitoring data provides an opportunity to validate radioecology models used to assess (14)C transfer to grassland ecosystems. This article compares and discusses the ability of two different models to reproduce the observed temporal variability in grass (14)C activity in the vicinity of AREVA-NC La Hague. These two models are the TOCATTA model which is specifically designed for modelling transfer of (14)C and tritium in the terrestrial environment, and PaSim, a pasture model for simulating grassland carbon and nitrogen cycling. Both TOCATTA and PaSim tend to under-estimate the magnitude of observed peaks in grass (14)C activity, although they reproduce the general trends. PaSim simulates (14)C activities in substrate and structural pools of the plant. We define a mean turn-over time for (14)C within the plant, which is based on both experimental data and the frequency of cuts. An adapted PaSim result is presented using the 15 and 20 day moving average results for the (14)C activity in the substrate pool, which shows a good match to the observations. This model reduces the Root Mean Square Error (RMSE) by nearly 40% in comparison to TOCATTA.

Sensitivity analysis for models of greenhouse gas emissions at farm level. Case study of N(2)O emissions simulated by the CERES-EGC model.

Modelling complex systems such as farms often requires quantification of a large number of input factors. Sensitivity analyses are useful to reduce the number of input factors that are required to be measured or estimated accurately. Three methods of sensitivity analysis (the Morris method, the rank regression and correlation method and the Extended Fourier Amplitude Sensitivity Test method) were compared in the case of the CERES-EGC model applied to crops of a dairy farm. The qualitative Morris method provided a screening of the input factors. The two other quantitative methods were used to investigate more thoroughly the effects of input factors on output variables. Despite differences in terms of concepts and assumptions, the three methods provided similar results. Among the 44 factors under study, N(2)O emissions were mainly sensitive to the fraction of N(2)O emitted during denitrification, the maximum rate of nitrification, the soil bulk density and the cropland area.