New insights into the association between lameness, behavior, and performance in Simmental cows

摘要

In a herd of 100 milking Simmental cows, data ofperformance and behavior parameters were collectedautomatically with different systems such as pedometers,an automatic milking system, and automaticweighing troughs for 1 yr. Performance measures wereseveral milking-related parameters, live weight, as wellas feed intake. Behavior-associated measures were feedingbehavior (e.g. feeding duration, number of visits tothe trough, and feeding pace) as well as activity suchas lying duration, number of lying bouts, and overallactivity. In the same time, lameness status of every cowwas assessed with weekly locomotion scoring. Accordingto the score animals were then classified lame (score4 or 5) or nonlame (score 1, 2, or 3). From these data intotal, 25 parameters summarized to daily values wereevaluated for their ability to determine the lamenessstatus of a cow. Data were analyzed with a regularizedregression method called elastic net with the outcomelame or nonlame. The final model had a high predictionaccuracy with an area under the curve of 0.91[95% confidence interval (CI) = 0.88–0.94]. Specificitywas 0.81 (95% CI = 0.73–0.85) and sensitivity was 0.94(95% CI = 0.88–1.00). The most important factors associatedwith a cow being lame were number of meals,average feed intake per meal, and average duration of ameal. Lame cows fed in fewer and shorter meals with adecreased intake per meal. Milk yield and lying-behavior-associated parameters were relevant in the model,too, but only as parts of interaction terms demonstratingtheir strong dependence on other factors. A highermilk yield only resulted in higher risk of being lame iffeed intake was decreased. The same accounts for lyingduration: only if lying time was below the 50% quantiledid an increased milk yield result in a higher risk ofbeing lame. The association of lameness and daily lyingduration was influenced by daily feeding duration andfeeding duration at daytime. The results of the studygive deeper insights on how the association betweenbehavior and performance parameters and lameness isinfluenced by intrinsic factors in particular and thatmany of these have to be considered when trying topredict lameness based on such data. The findings leadto a better understanding why, for instance, lying durationor milk yield seem to be highly correlated withlameness in cows but still have not been overly useful asparameters in other lameness detection models.