Housing pigs with undocked tails: methods for the early detection and prevention of tail biting
by Philipp Heseker
Date of Examination:2025-03-20
Date of issue:2025-04-15
Advisor:Prof. Dr. Imke Traulsen
Referee:Prof. Dr. Imke Traulsen
Referee:Prof. Dr. Jens Tetens
Persistent Address:
http://dx.doi.org/10.53846/goediss-11207
Files in this item
Name:Diss_PH_final.pdf
Size:1.79Mb
Format:PDF
Description:Dissertation
The following license files are associated with this item:
Abstract
English
The occurrence of tail biting is a widespread problem in global pig farming. Painful biting incidents lead to a significant impairment of animal health, animal welfare and the performance of pigs in various production stages. The background to this multifactorial behavioral disorder is very complex due to the large number of risk factors involved, which make targeted prevention and prediction almost impossible. Docking the pigs’ tails within the first few days of life has become a common and effective prevention strategy in agricultural practice, which is a short but painful procedure for the animal. This method has been officially banned in the EU for several years now, but it is still permitted if a farm has proven observed tail biting in its herd and other preventive measures have been implemented but have failed. Nevertheless, alternative strategies to prevent tail biting are needed in order to gradually avoid tail docking in the long term. The provision of organic enrichment material, for example, has proven to be particularly effective in promoting the animals' natural exploratory behavior and distracting them from the actual biting. In addition, the early detection and intervention of tail biting are essential in order to stop it as soon as the first biting occurred and to prevent it from escalating in the entire herd. For this approach, the potential and implementation of various digital systems is in the focus of the research and increasingly tested in practice. Those systems make it possible to continuously and objectively record animal behavior and housing conditions in order to individually support pig farms in their daily work. The aim of this dissertation was to investigate possibilities for the prevention and early detection of tail biting in weaner pigs with intact tails by implementing different sensor systems in a piglet rearing compartment. To prevent tail biting, an automatic enrichment device with pen-specific supplies was installed in the compartment to regularly provide the pigs with pelleted organic material. This serves not only to prevent tail biting, but also to support the animals' natural exploratory behavior and thus increase animal welfare, which is to be evaluated in this dissertation. To objectively observe the animals' behavior, cameras with connected microphones were installed above each pen for continuously recording. In addition to the existing and described early indicators for tail biting, new methods for the early detection of this behavioral disorder in weaner pigs should be researched. The first study investigated whether tail biting and in particular individual tail biters can be identified by recognizing pig screams in sound recordings, as the rapid removal of tail biters is an effective intervention strategy to maintain animal health and welfare. For this approach, 288 undocked rearing piglets were housed in six pens during two batches. The pigs were marked individually by colored ear tags and their tails were examined biweekly. The tail posture of the animals was recorded daily as a common early indicator for ongoing biting. As soon as the barn staff identified a tail biter (n = 7) during the daily animal checks, the video and audio recordings of the previous days were retrospectively analyzed for pig screams (sudden increase in volume with frequencies above 1 kHz) and evaluated in the video until no more biting incidents were observed anymore. A total of 2,893 screams were observed and analyzed in four tail-biting pens. Of these screams, 52.9 % were caused by tail biting in the observed pen, 25.6 % originated from other pens, 8.8 % were unclassifiable and 12.7 % were due to other causes. In the case of a tail biting incident, the screams were individually assigned to the biter and victim pigs by identifying them by their colored ear tags. Based on the audio analysis, tail biters were identified between one and nine days before they were removed by the farm staff. The screams were also recognized earlier than the increase in hanging tails and could therefore be favored as an early warning indicator. Analyzing the vocalizations of pigs to detect screams has the potential to identify tail biters at an early stage. In combination with individual animal markings and automatic analysis algorithms, tail biters could be identified and tail biting efficiently reduced. This would help the farmers to react faster and remove tail biters from the pen earlier in order to improve the health and welfare of the other animals. In the second study, the effects of automatically supplied enrichment material of three different types (alfalfa pellets, oat bran pellets or a mixture of both) and different enrichment frequencies (2, 4 or 6 supplies/day) on the behavior, tail biting incidence and daily weight gain of weaned piglets were investigated during six batches. The results showed significant effects and interactions of material, frequency and time of day on exploratory behavior, the incidence of tail biting and daily weight gain. Pigs were most likely to use the enrichment material when they received only two supplies per day or oat bran pellets. In addition, more pigs explored the material when it was provided in the afternoon compared to the morning. Tail lesions increased significantly from the fourth week of rearing. The probability of a tail lesion occurring was highest in groups with two supplies per day. The groups that received two supplies per day also had the highest probabilities of a tail loss at the end of the rearing period, with 0.170 for alfalfa pellets, 0.342 for the mixture and 0.486 for oat bran pellets. For daily weight gain, only the alfalfa groups differed significantly from the mixture groups at two supplies/day. There were no differences for the other factor combinations. These results show the potential of an automated enrichment device to support the pigs in performing their natural exploratory behavior in a conventional housing system. A higher number of daily supplies is beneficial for reducing the occurrence of tail lesions and tail losses. The third study builds on the first study with the aim of automatically detecting pig screams for the early identification of tail biters by developing a detector software that automatically analyses video and audio recordings and detects screams that can be caused by tail biting incidents. For this purpose, video recordings of rearing piglets from different pens and two batches were recorded over a period of 39 days each. Screams were automatically identified by the developed detector by analyzing the audio track of the video recordings and then evaluated by the outcome from the two batches of the first study. The results of the detector were assessed by testing various user-defined parameters (e.g. volume threshold, call duration, frequency range). A total of 2,898 screams were detected manually in four pens (748 hours), with 52.8 % caused by tail biting. In the second step, 89.9% of the detected screams were correctly classified as screams by the detector (accuracy 0.81-1.00), of which 64.0% were caused by tail biting. The detector automatically analyses the latest videos in user-defined time periods (e.g. hourly) and sends alerts via email if a threshold of screams per hour is exceeded in a pen. Subsequently, the detector extracts video clips that can be used for visually identifying the tail biters. In summary, the automatic analysis of pig screams has great potential for early detection and intervention of tail biting by recognizing tail biting events directly through the screams and providing video clips for identification of tail biters. All in all, the various systems in the piglet rearing compartment had a positive influence and high potential on the prevention and early detection of tail biting. The automatic enrichment device was able to reduce the risk of tail injuries by increasing the enrichment frequency and encouraging the pigs’ natural exploratory behavior. Targeted evaluation of video recordings with an integrated audio track enables automatic scream detection, which can be used for identifying tail biting incidents. The direct detection of biting incidents in particular offers major advantages compared to other early indicators, as targeted intervention measures can be initiated like identifying and removing the biter. The automatic scream detection has great potential to support pig farms in their daily work and has already been tested over several rounds under practical conditions.
Keywords: animal husbandry; tail biting; precision livestock farming; enrichment; weaner pigs
