Confinement to farrowing crates is known to prevent sows from performing natural behavior, impairing animal welfare and possibly causing chronic stress. Hair cortisol analyses are increasingly used to detect chronic stress in animals. In the present study, hair samples were collected in the neck of sows kept either in farrowing crates (FC, n = 31) or in a loose-housing system (LH, n = 30) in six batches. Cortisol was extracted and analyzed using chemiluminescence immunoassay. Mean hair cortisol concentrations (HCC) did not differ significantly between the systems (LH: 1.85 ± 0.82 pg/mg, FC: 2.13 ± 1.53 pg/mg, P = 0.631). HCC was also not affected by other factors, such as sows' parity, number of piglets, skin lesion score or sow's weight loss during the farrowing period. However, highly significant differences were found in hair growth rates between different regions within the 20 × 30 cm shaving area. While the hair in both lateral parts of the shaving area grew almost identically (left: 7.48 ± 3.52 mm, right: 7.44 ± 3.24 mm, P = 1.00), the hair grew more in the area above the spine (12.27 + 3.95 mm, P < 0.001). In both systems, the mean individual lesion score of sows declined from the beginning to the end of the housing period (P < 0.001). No difference was found between FC and LH sows at any time (P > 0.05). Since neither the amount of skin lesions nor HCC differed between LH and FC sows, it may be concluded that confining sows in farrowing crates did not affect chronic stress levels. However, results may be affected by a downregulation of the hypothalamic-pituitary-adrenal axis during long-term stress, resulting in lower cortisol levels over time. HCC in sows may also be influenced by a dominant stressor, such as farrowing or the presence of suckling piglets. Thus, for a comparison of different farrowing systems regarding chronic stress, the use of hair cortisol measurement seems to be limited. The present results revealed that differences in hair growth rate within the same body region exist. This important finding should be considered when collecting hair samples in pigs, since hair cortisol concentrations may vary depending on hair growth and length.

Sows confined to farrowing crates are restricted in performing natural behaviour such as maternal behaviour. Loose-housing farrowing pens (LH) and farrowing pens with crates (FC) were compared regarding sows’ nursing behaviour via video analyses over four weeks per batch (one day per week). Nursing frequency was similar in LH and FC pens (1.25 ± 0.82 vs. 1.19 ± 0.75 nursings/sow/hour; p > 0.05). However, nursing duration differed between the two systems (LH: 5.7 ± 4.6 min vs. FC: 7.0 ± 5.0 min; odds ratio (OR) 1.168, p = 0.011). In LH pens, more nursing bouts were sow-terminated than in FC pens (OR 0.427, p = 0.001). The probability of sow-terminated nursing occurring increased from week 1 to week 4 (OR 3.479, adjusted p (padj) < 0.001), while that of observing unnursed piglets decreased from week 1 to week 4 (OR 0.301, padj < 0.001) and rose with increasing litter size (OR 1.174, p = 0.010). We conclude that nursing behaviour was affected by the farrowing system, with shorter nursing duration and more nursing terminations by the sow in LH than in FC pens. Since this corresponds to the nursing behaviour of sows in semi-natural conditions, it can be assumed that sows in LH pens are more likely to exhibit natural nursing behaviour.