Calving prediction in dairy cattle by visual observation or the use of technical devices | Vorhersage der Abkalbung bei Milchkühen durch visuelle Beobachtung oder den Einsatz technischer Hilfsmittel

The overall objectives of this thesis were (1) to analyze the predictive performance of a calving sensor that recently became available, (2) to assess the tolerance of the sensor attachment at the animal´s tail and (3) to elicit current recommendations on calving management from SMEs and non-SMEs in order to provide farmers with best practice advice for managing the periparturient cow. In our first study, a total of 180 clinically healthy animals (157 multiparous, 23 primiparous) were enrolled. They were equipped with a calving sensor (Moocall calving sensor, Moocall Ltd., Dublin, Ireland) as soon as their parturition score (PS) reached the threshold of four. A cow with a PS of < 4 was defined as being more than 12 hours before parturition (Streyl et al., 2019). Enrolled animals were observed hourly for 24 h/d until calving occurred or the animal was excluded for other reasons. Investigators checked for signs indicating the onset of stage II of parturition, verified the position of the sensor and evaluated skin integrity of the tail above and below the sensor. Sensitivity and specificity of the tail-mounted calving sensor were analyzed for intervals of 1, 2, 4, 12, and 24 hours before the onset of stage II of parturition. Visual observation of the amniotic sac or feet outside the vulva were considered as gold standard and used to calculate sensitivity and specificity. A calving alarm is generated through increased tail raising frequency before parturition and is distinguished in two distinct alarm types: HTA1h (i.e., high tail activity in the last hour) and HTA2h (i.e., high tail activity in the last 2 hours). The sensitivity of both HTA1h- and HTA2h-alarms was 19% for 1-hour-interval before onset of stage II of parturition. Sensitivity for HTA1h- and HTA2h-alarms increased to 43%, and 39%, respectively for the 2-hour-interval and increased up to 75%, and 53%, respectively when broadening the interval to 24-hours before the onset of stage II of parturition. The position of the sensor was checked hourly and detachment events were evaluated. In 25 animals (13.9%), a sensor remained continuously attached to the tail until calving began. In 93 animals (51.7%), the sensors had to be reattached repeatedly until calving occurred. Overall, the sensors fell off 737 times during the study period. Here, heifers were significantly less likely to experience sensor loss (P = 0.022) or slipping (P = 0.012) but more likely to experience swelling or pain on the tail (P = 0.009). Skin integrity was assessed after each detachment event and was found to be intact in 206 events. Mild pressure marks were noted in 239 events, whereas severe pressure marks with swelling occurred after 180 events. Open wounds with/without swelling were detected in 6 detachment events and 6 others were assessed as necrosis. A sensitivity of 19% for detecting impending calving 1 h before the onset of the second stage of parturition is very low and therefore not recommendable as on-farm device. Further work is needed to develop a lighter and a more reliable and less invasive mounting device to improve attachment and skin compatibility. In the second study, subject matter experts (SMEs), who had published on calving management and veterinary practioners (non-SMEs) were asked to participate in a survey about calving management. A total of 104 participants (80 SMEs via online survey and 24 non- SMEs via survey on a workshop) were contacted, with response rates of 67.5% and 100%, respectively. Only 38 participants (36.5%) complete questionnaires were returned. Data from incomplete questionnaires were also included and adjusted to the number of participants per question. The two groups of veterinarians differed little in their opinion of calving management. The distinction between the two stages of parturition was important, and for each stage they recommended watching for signs of impending parturition, such as "Restlessness," "Tail raising", "Relaxation of pelvic ligaments," and "Vaginal discharge" for stage I and "Visibility of fetal parts", "Visible amniotic sac" and "Abdominal contractions" for stage II of parturition. The authors of other publications assigned these signs to many different periods before parturition, and it became clear that the imprecision in defining the signs (e.g., restlessness) made accurate temporal assignment impossible. The widely varying recommendations from study participants on when to begin monitoring pregnant cows ranged from one to 21 days prior to calving date. The literature recommends moving pregnant cows 21 to 14 days before the estimated calving date until the start of the second stage (Cook et al., 2007), with observation intervals of approximately every 3 to 6 hours for cows in the first stage of parturition (Mee, 2004). Most study participants recommended longer intervals of 12 or 6 hours, again consistent with the results of farmer surveys (Vasseur et al., 2010; Santos and Bittar, 2015; Cummins et al., 2016; Villettaz- Robichaud et al., 2016). Once cows enter the second stage of parturition, they should be moved to a maternity pen (Cook et al., 2007) where they should be observed continuously or every 30 min (Mee, 2004). For cows already housed in a maternity barn, most participants in our study recommended an observation interval of every 6 hours or every 2 hours. Since many calvings occur at night, these observation intervals may be inappropriate (van Keyserlingk and Weary, 2007), but it needs to be mentioned here that the stage of parturition was not defined in that question, so participants may have been misled. To detect the onset of parturition technical calving devices were not unanimously approved by study participants (55.3%), but vaginally inserted temperature loggers were most frequently recommended. Authors of several publications on sensor technology have summarized that current sensors are not suitable as a sole tool for calving detection (Aoki et al., 2005; Burfeind et al., 2011; Palombi et al., 2013; Ouellet et al., 2016; Henningsen et al., 2017; Giarette et al., 2021). Most study participants (62.2%) and the current literature underlined the major impact of calving personnel expertise on the timing of cow movement prior to calving. Sixteen participants (35.6%) fully agreed, that the timing of cow movement relative to the stage of parturition has an influence on the calving process, supporting study results of Carrier et al. (2006) and Proudfoot et al. (2013). In some studies, cows were given a vaginal examination if no progress of parturition was observed 2 h after AS burst (Wehrend et al., 2006) or feet were visible (Johanson et al., 2003). Most study participants recommended a vaginal examination within 1 h after observing the AS or feet outside the vulva. Overall, this thesis showed that (1) the sensor investigated had a poor sensitivity and, in addition to frequent sensor slipping and sensor losses, the tail-mounted calving sensor caused pressure marks and necrosis and further research is required, (2) the survey showed good consensus between SMEs and non-SMEs, but information found in the literature lacked high clinical utility due to imprecision in either definition, measurement, timing or their poor evidential base, and (3) there is broad agreement on how to identify imminent signs of parturition, but more research is warranted to determine the best time for moving cows to the maternity pen.