Pig tail biting in different farrowing and rearing systems with a focus on tail lesions, tail losses and activity monitoring
von Maria Gentz
Datum der mündl. Prüfung:2020-07-09
Betreuer:Prof. Dr. Imke Traulsen
Gutachter:Prof. Dr. Jürgen Hummel
Gutachter:Prof. Dr. Eva Schlecht
EnglischTail biting is a widespread behavioural disorder in modern pig housing systems. It is defined as an oral manipulation of a pig’s tail by another. Different types of tail biting are described throughout the literature which are discernible by the way of biting as well as multifactorial causes. Due to the prohibition of routine tail docking in the EU, adapted housing systems for undocked pigs have gained importance. The aim of this dissertation is to investigate the effect of farrowing and rearing systems on tail lesions and losses of pigs, with a focus on undocked animals. Different housing systems, as described below, are taken into consideration as a means to evaluate their influence on behaviour abnormalities during rearing and fattening periods. In Paper One, docked and undocked pigs are compared regarding their tail lesions and losses during rearing and fattening periods. The pigs are housed in three farrowing systems: conventional farrowing crates (FC); free farrowing pens (FF); and group housing of lactating sows (GH). Additionally, they are divided into two rearing systems: a conventional system (CONV), where the pigs were regrouped and transferred to conventional finishing pens at ten weeks of age; or a wean-to-finish (W-F) system, where the pigs remained in their pens until slaughter with increased space allowance during rearing. Significant differences in the rearing systems are observed through the weekly assessment of tail lesions and losses. Fewer lesions occur for the W-F pigs (maximum: CONV 58.01 %, W-F 41.16 %). Docked animals also experience statistically significantly fewer lesions than undocked animals. The W-F reared pigs have significantly fewer tail losses, as recorded at the end of fattening period (CONV 67.63 %, W-F 38.2 %). In conclusion, a long-term effect of early socialisation does not appear to influence tail biting, given that farrowing system does not have a statistically significant effect on tail lesions. Nonetheless, an increased space allowance and reduced regroupings included in the W-F rearing system decreases the frequency of tail lesions and losses, significantly. Due to the findings of paper one, pigs with tail lesions were classified and characterised based on the frequency and duration of the lesion to create a latent parameter. Paper Two investigates whether pigs with different severity of tail lesions can be separated into biologically relevant groups using a cluster analysis. Pigs housed in FC, FF, GH and reared in CONV and W-F are included in the analysis. Animals are indicated according to five lesion groups: from (I) no lesions to (V) many long lasting lesions; the separability of the predefined lesion groups is checked by an animal individual lesion parameter. As docking status is the main parameter affecting tail lesions, it is also included in the analysis. Using a k-means cluster analysis, groups I and V, are clearly distinguishable; however, the middle lesion groups are less defined within the analysis. In conclusion, not all pigs are affected by tail biting to the same extent. Docking status as well as increased space allowance can lead to a reduction in tail lesions. Paper Three examines the long-term effects of artificial rearing (AR) on tail biting and development of pigs. AR and FC pigs are compared with a focus on tail lesions and losses, skin lesions, and performance traits. In this study, all pigs are raised in a conventional rearing system. Results show that AR pigs exhibit fewer lesions than the FC pigs; less than 15 % of AR pigs had tail lesions. At the end of fattening period, only 6.9 % of the AR pigs had lost parts of their tails, which was significantly different from FC pigs. AR pigs also have significantly fewer skin lesions after rehousing to rearing. After regrouping and rehousing to fattening, skin lesion frequency is not differing. Performance traits differ during nursery period in favour of the FC pigs, but in contrast, AR pigs have the higher daily weight gains during rearing period. Fattening period does not have a significant effect. AR pigs exhibit fewer tail lesions and losses with comparable daily weight gains, which is a result of reduced regroupings and identical housing conditions during rearing and fattening periods. However, these results should be critically reflected in terms of animal welfare and economic feasibility, prior to making practical recommendations. Paper Four focusses on the activity behaviour of pigs. Animals’ behaviours are recorded and evaluated manually and automatically at three time points during the rearing period (beginning, middle, end). The manual analysis uses the point sampling method and determines the respective activity of every pig every 12 minutes. The automatic evaluation is based on a computer vision algorithm and records the pigs as ellipses. The coordinates as well as the centers of the individual animal ellipses are measured. By linking the nearest neighbour ellipses, the movement of the pigs is recorded within 50 frames, adjusted to the manual time samples. The evaluation is carried out at pen level. The results show that the manual animal detection is more precise, but the automated analysis has a high accuracy and is only limited by objects that obscure the pigs. Humans can interpret the animals’ behaviour, even though single body parts are not visible, whereas automatic analysis systems have difficulties in adapting this human knowledge about the shape and movements of animal bodies. Nonetheless, the daily course of the manually and automatically assessed activity and the trends of increasing or decreasing activity are comparable. Furthermore, days can be automatically divided into inactive, active and very active phases with a high accuracy. Overall, there is a high level of agreement between the two methods of analysis. In summary, it can be concluded that the housing system plays a particularly important role during the rearing period in influencing tail biting behaviour, with long-term effects on tail lesions and losses. Increased space allowance and reduced number of regroupings, as with the W-F pigs, appears to have a positive effect in reducing lesions. There remains a large difference in the number and severity of tail lesions and losses between docked and undocked pigs. Characteristics of pigs with and without tail lesions are identifiable using a cluster analysis. Despite animal welfare concerns, artificially reared pigs benefit from maintaining stable groups and exhibit fewer tail lesions and losses, less skin lesions, and their performance is comparable to pigs reared more traditionally. Fewer regroupings, more space during rearing and therefore less stress due to not needing to form new hierarchies contribute to less tail biting. Furthermore, the automatic behaviour analysis has shown to be able to successfully separate active and inactive time intervals at a group level. It promises a high potential, especially regarding early warning indicators for tail biting.
Keywords: tail lesions; tail losses; farrowing system; rearing system; undocked pigs; docked pigs; rearing; fattening; scoring; assessment; stress; classification; artificial rearing; reduced regrouping; Activity; algorithm; nearest neighbour; activity index; behaviour analysis; distance index