Deviation of behavioural and productive parameters in dairy cows due to a lameness event: a synthesis of reviews

Lameness is a widespread multifactorial condition affecting the health and performance of dairy cows. Despite the growing support by precision farming technologies, farmers still lack reliable data-driven tools to early identify lame cows. This study used a synthesis of reviews to identify cow’s behavioural and productive parameters most related to lameness and estimate their deviation due to a lameness event. The methodological approach used reviews as starting point to identify the most pertinent studies with the intention of extracting and analysing data from these primary studies. The final dataset used information collected from 31 research papers, cited in 15 reviews, and involved more than 25,000 dairy cows. Five parameters were suitable for the meta-analysis: one about eating behaviour (eating time), three regarding activity and resting behaviour (lying bouts, lying bout duration and lying time) and milk yield. The meta-analysis revealed that all parameters had a significant deviation in cows affected by lameness. The calculation of the pooled means allowed to quantify a mean value for the deviation imposed by a severe lameness event from the value recorded on nonlame cows. Compared to a nonlame animal, a lame cow had a significant negative deviation for eating time (−39 min/day), number of lying bouts (−0.5/day), and milk yield (−3 kg/day). Lame cows had positive deviations for lying bout duration (+12 min/bout) and daily lying time (+42 min/day). The individual or combined use of these mean deviation values as alarm reference thresholds could improve the accuracy of the current automated lameness detection systems. Feeding and resting habits, and milk yield had a significant deviation in lame cows.Lame cows decreased their time spent eating, lying bouts frequency, and milk yield.Lame cows had prolonged lying bout duration and daily lying time. Feeding and resting habits, and milk yield had a significant deviation in lame cows. Lame cows decreased their time spent eating, lying bouts frequency, and milk yield. Lame cows had prolonged lying bout duration and daily lying time.跛行是一种广泛流行的多因素疾病，会损害奶牛的健康与生产性能。尽管精准农业（precision farming）技术的应用支持不断增强，但养殖户仍缺乏可靠的数据驱动型工具，以实现奶牛跛行的早期识别。本研究采用综述整合的研究策略，筛选出与奶牛跛行关联性最强的行为与生产参数，并评估跛行事件引发的参数偏差。本研究的方法论以综述文献为起点，筛选出最相关的原始研究，进而提取并分析这些原始研究中的数据。最终构建的数据集涵盖了来自31篇研究论文的信息，这些论文被15篇综述文献引用，涉及超过25000头奶牛。共有5项参数适用于荟萃分析（meta-analysis）：1项与进食行为相关，即进食时长（eating time）；3项与活动和休息行为相关，即躺卧频次（lying bouts）、单次躺卧时长（lying bout duration）以及每日躺卧时长（lying time）；以及产奶量（milk yield）。荟萃分析结果显示，跛行奶牛的所有上述参数均存在显著偏差。通过计算合并均值，本研究量化了严重跛行事件相较于健康奶牛的参数偏差平均值。与健康奶牛相比，跛行奶牛的进食时长（-39分钟/天）、躺卧频次（-0.5次/天）以及产奶量（-3千克/天）均出现显著负向偏差。而跛行奶牛的单次躺卧时长（+12分钟/次）与每日躺卧时长（+42分钟/天）则呈现显著正向偏差。将这些平均偏差值单独或组合用作报警参考阈值，可提升现有自动化跛行检测系统的识别精度。进食与休息习惯以及产奶量在跛行奶牛中均存在显著偏差。跛行奶牛的进食时长、躺卧频次以及产奶量均有所下降。跛行奶牛的单次躺卧时长与每日躺卧时长均有所延长。进食与休息习惯以及产奶量在跛行奶牛中均存在显著偏差。跛行奶牛的进食时长、躺卧频次以及产奶量均有所下降。跛行奶牛的单次躺卧时长与每日躺卧时长均有所延长。