After the biennial Ruminant Lameness Conference (held recently in Venice), it is always a good time to reassess our position in lameness control and management (and our future direction). The combination of invited plenary speakers and selected oral presentations usually provides an overview of where we have been and where we are going, and the Venice conference was no exception.
It is important to start with the conclusion from a recent review (Thomsen and Shearer, 2023): despite success at the individual farm level, overall lameness prevalence has not changed in the last 35 years. This is despite the fact that we know far more about lameness treatment, management, control and prevention than we did in 1989. Lameness remains the most intractable welfare problem on dairy farms and a continued major source of reduced productivity and economic loss. It is clear that despite significant increases in our understanding of the problem over the last 40 years, we are nowhere near a solution.
We know that the system is the problem. Housing dairy cows on concrete causes lameness, while keeping cows permanently at pasture markedly reduces it. However, changing the system is not feasible. We need to reduce lameness within the system. This will require new knowledge, new approaches and a commitment to working with farmers and their advisers. We need a paradigm shift – the status quo is not working.
Rather than searching for simple silver bullets (‘give cows more access to pasture,’ ‘treat all cows with non-steroidal,’ ‘feed more zinc’), we need to combine strategies to identify effective programs that will result in less lameness on dairy farms. Attending the lameness conference identified that such an approach is where we need to head and that recent advances in technology and knowledge mean that it is increasingly feasible to combine strategies. The key problem is that testing this approach will require large-scale data collection and, thus, large-scale funding. Without this, we will likely continue as we have, with too many lame cows on too many farms. The key developments are automatic lameness detection, a renewed focus on genetics, and a better understanding of what producers want.
Automatic lameness detection has been available since the late 1990s but has never been accurate enough for use on commercial dairy farms. However, recent advances in AI-enhanced computer vision methods mean that automated lameness detection is now much closer to commercialisation, with systems available in many countries. These systems are more accurate than older systems but still have the issue that their sensitivity and specificity are <100%. The sensitivity is not an issue, as automatic monitoring means monitoring every time a cow walks past the camera. However, the less-than-perfect specificity means that it identifies as lame cows that are not lame, which can significantly impact farmer buy-in and approach. This is a greatly underestimated problem, particularly in non-veterinary journals.
Some papers claim that new systems are highly accurate, but on closer inspection, their specificity is around 90%. This sounds very high, but it means that for every 100 non-lame cows, the system will identify 10 cows as lame. That is acceptable for mastitis, as the check (udder palpation and stripping milk into a California Mastitis Test) is quick and easy. However, getting a cow into a crush and examining the feet when it is a false positive quickly becomes annoying and wastes time. We need a set of target standards for commercially available systems so that farmers can confidently detect lameness. These need to be independently monitored and not just based on ‘company data’.
Nonetheless, automatic lameness detection provides a huge amount of information that can be used for purposes other than simply detecting lame cows. If collected on a large scale over multiple farms, this data can give us new avenues to explore, particularly in relation to genetics. Genetics is another recent advance area that will be integral to any multi-pronged strategy to reduce lameness. In the past, research on the genetics of lameness has had a limited impact on lameness outcomes, principally because of its focus on easily measurable but indirect predictors (such as leg conformation) and generally low heritabilities. However, with the increasing recognition that selection on direct health traits can accelerate genetic gain, there is increasing belief that focusing on the genetics of lameness could be crucial in reducing lameness impact in dairy cows.
One focus of ‘direct’ genetic assessment has been lesion data collected during routine hoof trimming. With hoof trimming now established as a key means of reducing lameness prevalence at the herd level, and with effective, accurate lesion recording now being more common, we need large-scale studies to show the value of using lesion data to guide genetic selection. However, direct genetic assessment using farm lameness records may be even more beneficial. The problem with lameness records is their quality – what proportion of lame cows are detected, how accurate is lameness detection, and what proportion of lame cows are actually recorded? Automatic lameness detection solves these problems while also providing far more data, such as how long the cow was lame, how severe it was, and how it progressed through the stages of lameness. Again, properly designed large-scale studies are needed to assess the value of automatic lameness detection data in developing a genetic lameness index. As farmer lameness records have already been used to create such indices, it seems highly likely that the repeated nature of automatically collected data will be even more effective.
Farmer resistance to the use of automatic detectors is likely to play a major role in the value of such research. However, if successful, this research is likely to significantly increase the uptake of automatic detectors, which will have flow-on effects on the detection of lameness in individual cows and, hopefully, lead to earlier and more effective treatment. This highlights that, irrespective of the research findings, the critical element in all this is buy-in from the producer. To achieve this, we must involve on-farm stakeholders from the start and build farmer participation at all stages. This is a critical social science and will need to use techniques such as participatory research and create a framework for taking the research outcomes onto farms and testing their impact and feasibility. This sounds simple but is perhaps the most difficult part of the research. One key issue is the region or country-specific nature of the on-farm stakeholders and their different aims, goals, strengths and weaknesses. Nevertheless, it is important not to over-emphasise these differences – recent research has shown that farmers are concerned about lameness irrespective of their herd prevalence, even when low. What is needed is effective, feasible methods of lameness control rather than a strategy to change farmer opinion.
Putting multiple strategies together can radically alter our approach to lameness management and lameness outcomes. Combining improving technology with large-scale datasets and genetic analysis alongside a farmer focus should lead to the development of new tools for use on farms that can be adopted by dairy farms across the world, favourably impacting economics, greenhouse gas production, and animal welfare. Maintaining the status quo is not good enough.