Behavioural and physiological consequences of tooth resection in commercial piglets: implications for welfare

BACKGROUND: In commercial farming the shortening of piglet needle teeth is a common but controversial and legally restricted practice due to its potential to induce pain and stress. However, very limited data on the long term behavioural and physiological effects exist. AIM: To investigate both short- and long-term behavioural and physiological indicators of pain and / or stress in piglets subjected to tooth resection by the two commercially employed methods (clipping and grinding) and to compare these methods in terms of piglet welfare. METHODS: Experiment (expt) 1: 120 piglets from 20 litter-based replicates (12 mixed sex replicates: n = 3 / sex / litter and 8 female-only replicates: n = 6 / litter) were weighed at birth and assigned to 1 of 3 treatments: 1) tooth clipping with sterilised pliers (Clip), 2) tooth grinding with a hand-held rotating grindstone (Grind), 3) sham grinding (Sham). On treatment day (d1) piglets were filmed for 1 minute before and after treatment. Treatment duration, remaining tooth lengths and presence of blood were recorded. Pre- and post-treatment behaviours were blind-scored from the recordings. Expt 2: 68 piglets from expt 1 (34 males and 34 females from 12 litters; n = 1 / sex / treatment / litter (4 piglets excluded)) were remotely filmed within the home pen for 2 hour periods on d1, d5, d12, d26, d33 and d40. Recorded behaviours and teat order were blind-scored using continuous focal observation. Piglet weight, tear staining, and face and body lesions were recorded weekly (d6, d13, d20, d27, d34, and d41). Expt 3: Post-mortem (PM) tissue samples (needle teeth and trigeminal ganglia (TG)) were collected from 48 female piglets from expt 1 (8 litters; n = 6 / litter) either 7 or 42 days after treatment (n = 1 / treatment / PM group) with blood samples taken 1 day prior. Haematological analyses, tooth damage scoring, and real time RT-qPCR analysis of gene expression in the tooth pulp (TP) and TG were conducted. RESULTS: Expt 1: Compared to sham, mean remaining needle tooth lengths were 0.9 mm and 1.4 mm shorter in the grind and clip groups respectively (p < 0.001), and (minor) bleeding was observed in 97.5 % and 22.5 % of clip and grind piglets. Resection treatment durations did not differ and handling in all treatments caused stress and / or discomfort as evidenced by reductions in locomotion and exploratory behaviour (p < 0.001), an increase in head flicks (p < 0.001) and in time spent with alert (forward) ear positioning (p < 0.01). Treatment induced champing (opening and closing the jaws) in 80 % of clipped piglets but only 45 % of sham piglets (p < 0.01) with grind piglets intermediate (60 %, p > 0.05 vs. clip and sham). Sham piglets investigated the wood shavings more after treatment (p < 0.01) but this behaviour was unchanged in resected piglets. Expt 2: Overall weight gain, skin lesions, teat order, and tear staining were unaffected by either resection method up to d42 (p > 0.05). Behaviours changed over time and aggressive behaviours were more apparent in males from as early as 2 weeks of age but only one behaviour was notably altered by treatment. Head swipes / bites were observed at a greater frequency in sham vs. clip piglets (11.9 vs 7.9, p < 0.05) with grind intermediate (10.8, p > 0.05 vs. clip and sham). Expt 3: Blood composition and plasma concentrations of cortisol and haptoglobin were unaffected by treatment. Clipped teeth had a greater frequency and severity of visible tooth damage (clip wTDI: 18.34; grind wTDI: 12.76, p < 0.001). Despite this, relative to the sham group, TP gene expression of inflammatory markers was upregulated to a similar extent in both resection groups (CXCL8 gene fold change (FC): 331 and 519; PTGS2 gene FC: 1.8 and 2.2 for ground and clipped teeth respectively, (p > 0.05)). However, expression of PTGER2 was only significantly upregulated in clipped teeth (Clip FC: 1.39, p < 0.01, Grind FC: 1.17, p > 0.05 vs. clip and sham),CALCB (involved in pain signalling and reparative dentine formation) was downregulated differentially (grind FC: 0.5, clip FC: 0.2, p < 0.001), and upregulation of GFAP (a marker of nerve damage and potentially nociceptor sensitisation) in the TG was only significantly higher than sham expression in the grind group (Grind FC: 1.4 (p < 0.05 vs. sham); clip FC: 1.2 (p > 0.05 vs grind and sham)). CONCLUSIONS: Handling stress likely overwhelmed pain behaviours immediately after tooth resection though some limited and subtle behavioural indicators of pain were detected. Macroscopic tooth damage appeared worse after clipping, however, both methods caused chronic, and likely painful, pulpitis. Infections had not become systemic by d42 which was reflected in the lack of overt pain behaviours observed within the home pen, although grinding may have caused some level of neuropathy from d7. More intricate behavioural analyses, different tissue sampling intervals, and a wider range of physiological measures are required to confirm the findings of this study.