Changes in the drinking behaviour of pigs may indicate health, welfare or productivity problems. Automated monitoring and analysis of drinking behaviour could allow problems to be detected, thus improving farm productivity. A high frequency radio frequency identification (HF RFID) system was designed to register the drinking behaviour of individual pigs. HF RFID antennas were placed around four nipple drinkers and connected to a reader via a multiplexer. A total of 55 growing-finishing pigs were fitted with radio frequency identification (RFID) ear tags, one in each ear. RFID-based drinking visits were created from the RFID registrations using a bout criterion and a minimum and maximum duration criterion. The HF RFID system was successfully validated by comparing RFID-based visits with visual observations and flow meter measurements based on visit overlap. Sensitivity was at least 92%, specificity 93%, precision 90% and accuracy 93%. RFID-based drinking duration had a high correlation with observed drinking duration (R 2=0.88) and water usage (R 2=0.71). The number of registrations after applying the visit criteria had an even higher correlation with the same two variables (R 2=0.90 and 0.75, respectively). There was also a correlation between number of RFID visits and number of observed visits (R 2=0.84). The system provides good quality information about the drinking behaviour of individual pigs. As health or other problems affect the pigs' drinking behaviour, analysis of the RFID data could allow problems to be detected and signalled to the farmer. This information can help to improve the productivity and economics of the farm as well as the health and welfare of the pigs.

A dairy cow's lifetime resilience and her ability to recalve gain importance on dairy farms, as they affect all aspects of the sustainability of the dairy industry. Many modern farms today have milk meters and activity sensors that accurately measure yield and activity at a high frequency for monitoring purposes. We hypothesized that these same sensors can be used for precision phenotyping of complex traits such as lifetime resilience or productive life span. The objective of this study was to investigate whether lifetime resilience and productive life span of dairy cows can be predicted using sensor-derived proxies of first-parity sensor data. We used a data set from 27 Belgian and British dairy farms with an automated milking system containing at least 5 yr of successive measurements. All of these farms had milk meter data available, and 13 of these farms were also equipped with activity sensors. This subset was used to investigate the added value of activity meters to improve the model's prediction accuracy. To rank cows for lifetime resilience, a score was attributed to each cow based on her number of calvings, her 305-d milk yield, her age at first calving, her calving intervals, and the DIM at the moment of culling, taking her entire lifetime into account. Next, this lifetime resilience score was used to rank the cows within their herd, resulting in a lifetime resilience ranking. Based on this ranking, cows were classified in a low (last third), moderate (middle third), or high (first third) resilience category within farm. In total, 45 biologically sound sensor features were defined from the time series data, including measures of variability, lactation curve shape, milk yield perturbations, activity spikes indicating estrous events, and activity dynamics representing health events (e.g., drops in daily activity). These features, calculated on first-lactation data, were used to predict the lifetime resilience rank and, thus, to predict the classification within the herd (low, moderate, or high). Using a specific linear regression model progressively including features stepwise selected at farm level (cutoff P-value of 0.2), classification performances were between 35.9 and 70.0% (46.7 ± 8.0, mean ± SD) for milk yield features only, and between 46.7 and 84.0% (55.5 ± 12.1, mean ± SD) for lactation and activity features together. This is, respectively, 13.7 and 22.2% higher than what random classification would give. Moreover, using these individual farm models, only 3.5 and 2.3% of cows were classified high when they were actually low, or vice versa, whereas respectively 91.8 and 94.1% of wrongly classified animals were predicted in an adjacent category. The sensor features retained in the prediction equation of the individual farms differed across farms, which demonstrates the variability in culling and management strategies across farms and within farms over time. This lack of a common model structure across farms suggests the need to consider local (and evidence-based) culling management rules when developing decision support tools for dairy farms. With this study we showed the potential of precision phenotyping of complex traits based on biologically meaningful features derived from readily available sensor data. We conclude that first-lactation milk and activity sensor data have the potential to predict cows' lifetime resilience rankings within farms but that consistency between farms is currently lacking.

Milk yield dynamics during perturbations reflect how cows respond to challenges. This study investigated the characteristics of 62,406 perturbations from 16,604 lactation curves of dairy cows milked with an automated milking system at 50 Belgian, Dutch, and English farms. The unperturbed lactation curve representing the theoretical milk yield dynamics was estimated with an iterative procedure fitting a model on the daily milk yield data that was not part of a perturbation. Perturbations were defined as periods of at least 5 d of negative residuals having at least 1 day that the total daily milk production was below 80% of the estimated unperturbed lactation curve. Every perturbation was characterized and split in a development and a recovery phase. Based hereon, we calculated both the characteristics of the perturbation as a whole, and the duration, slopes, and milk losses in the phases separately. A 2-way ANOVA followed by a pairwise comparison of group means was carried out to detect differences between these characteristics in different lactation stages (early, mid-early, mid-late, and late) and parities (first, second, and third or higher). On average, 3.8 ± 1.9 (mean ± standard deviation) perturbations were detected per lactation in the first 305 d after calving, corresponding to an estimated 92.1 ± 135.8 kg of milk loss. Only 1% of the lactations had no perturbations. On average, 2.3 kg of milk was lost per day in the development phase, while the recovery phase corresponded to an average increase in milk production of 1.5 kg/d, and these phases lasted an average of 10.1 and 11.6 d, respectively. Perturbation characteristics were significantly different across parity and lactation stage groups, and early and mid-early perturbations in higher parities were found to be more severe with faster development rates, slower recovery rates, and higher milk losses. The method to characterize perturbations can be used for precision phenotyping purposes that look into the response of cows to challenges or that monitor applications (e.g., to evaluate the development and recovery of diseases and how these are affected by preventive actions or treatments).