The Effects of Light Level From 1 to 7 Weeks of Age, and Stress at 16 Weeks, on Plumage Damage and Injurious Pecking in Isa Brown Pullets Reared for Free-Range Egg Production

Severe feather pecking (SFP) is an abnormal behaviour which reduces both hen welfare and the efficiency of egg production. While the specific causes are unknown, it is acknowledged to be multifactorial, complex and unpredictable (Rodenburg et al., 2013; Hartcher et al., 2016). The problem is compounded because the behaviour seems to spread rapidly via social learning (Zeltner et al., 2000) and is therefore difficult to control in large groups of hens.

A reliable method to reduce pecking problems in indoor housing systems is to maintain low light intensity. However, there are practical difficulties in free-range systems due to light entering the shed via the pop holes. The effect of low light level during rearing on the development of SFP was investigated by Kjaer and Vestergaard (1999), who reared chicks to 15 weeks in 30 vs. 3 lux. At 10 weeks, pullets reared in the brighter light performed three times more SFP than the 3 lux treatment. Cannibalism also tended to occur more in the laying period in the 30 lux treatment. This and other research (Jensen et al., 2006; Gilani et al., 2012) suggested the existence of a sensitive period in early life for the development of IP in chicks. The present experiment therefore, investigated whether maintaining chicks under low lux in early life would provide long-term benefits by reducing SFP in free-range hens.

A second proposed factor in the development of SFP is elevated stress. In industry, the transfer of pullets from the rear to the layer farm occurs at about 14-16 weeks of age, and this process combines a number of ‘stressors’ - transport, re-housing and re-grouping. El-Lethey et al. (2000, 2001) showed evidence for the involvement of stress and elevated corticosterone concentrations in the expression of SFP, while Bestman and Wagenaar (2003) reported that bringing hens onto the farm at about 16 weeks was associated with higher levels of SFP, compared to rearing them on the egg farm or relocation to the new farm by about 10 weeks. Thus, stress around relocation at 16 weeks may also be an underlying cause of SFP.

This experiment tested two hypotheses: (1) chicks housed under low light in early rearing would have better plumage condition and lower mortality from pecking injuries as adults, and (2) the combined stressors of transport, relocation and mixing at 16 weeks would result in poorer plumage condition and higher mortality from pecking injuries as adults.

A total of 880 day-old, non-beak-trimmed chicks (ISA Brown) were purchased from a hatchery in Tamworth, NSW, Australia, and transported to the research facility. Chicks were randomly distributed among 16 floor pens (2.0 m × 3.2 m) bedded with wood shavings. Shed temperature, photoperiod and illumination were managed according to the ISA Brown Management Guide (2010), and feed and water were available ad libitum. At regular intervals from day 6, all birds were individually weighed and inspected for plumage damage (PD) on seven body areas (head, neck, back, rump, tail, side and vent). PD was recorded as either absent or present for each area. In addition, at the completion of the experiment (41 weeks) the level of PD at each area was scored as 0 = nil damage; 1 = some feathers missing; 2 = about half of the feathers missing; and 3 = most feathers missing (Tauson et al., 2005). Hens that died were necropsied to identify cause of death, and injured hens were assessed and treated as necessary.

The experiment had a 2×2 factorial arrangement with replication. Factor 1 compared light intensity between 1 and 7 weeks of age (< 5 lux: Low vs. min. 40 lux: Normal). At 9 weeks, the 16 pens of pullets were carefully relocated to another shed, whilst remaining in their original pengroups. In the new shed, groups were allocated at random to pens within blocks, based on shed aspect (north vs. south). All pens measured 1.83 m × 3.25 m, had wood shavings on the floor and contained a feeder, drinker, perch unit and nest-box unit (closed). At 14 weeks, the number of pullets was reduced to 50 per pen and photoperiod was increased weekly to stimulate egg production by providing an additional 30-min light per day, until 15hL:9hD was reached in week 19. Nest boxes were opened at 15 weeks, and at 16 weeks. Factor 2: Transport, Relocation and Mixing (TRM) was imposed on eight pens of pullets. Pullets were placed in poultry crates (max. 8 per crate) and transported by motor vehicle for 35-40 minutes. Upon return to the shed, crates were unloaded and pullets placed in a new pen, in which they were mixed 50/50 with both familiar and unfamiliar pullets. The other 8 pens were not treated (NT), that is, did not experience TRM. In week 20, the pop-holes to the outdoor ranges were opened.

Differences in mean bird weight per pen were analysed using ANOVA (Genstat) at each assessment date, after blocking on side of the shed. Number of PD sites per hen at each assessment date, and the highest (worst) feather score recorded across the 7 areas per hen at 41 weeks, were analysed using GLMM with binomial distribution and a logit link function in Genstat. Light and TRM were fixed effects and block/pen were random effects. Due to apparent differences in PD of hens in pens on the north compared to south side of the shed, analyses were repeated with shed block as a fixed effect, and pen as a random effect. Differences in the number of mortalities (including hens removed with pecking injuries) per pen were analysed using survival analysis with censored data based on week (Genstat).

While mean bird weight per pen did not differ due to the main effects at 6 days and 8 weeks of age (pooled means 68 and 683 g, respectively), at 14 weeks Low light pullets tended to weigh less than Normal pullets (1.296 vs 1.338 kg, respectively; sed 0.021, P = 0.074). However at 18 weeks, two weeks after imposing the TRM treatment, TRM pullets were heavier (P = 0.026) than NT pullets (1.623 vs 1.570 kg, respectively; sed 0.021). By 24 weeks the difference was a weak tendency (1.865 vs 1.816 kg, respectively; sed 0.026, P = 0.08) and thereafter there were no differences due to the main effects or interactions on average hen weight per pen.

PD was not apparent at, or prior to, 18 weeks. However, an outbreak of SFP/PD occurred soon after 18 weeks and, at the next assessment (24 weeks), ~40% of hens had at least some PD. Figure 1 shows the proportion of the flock that displayed PD from 18 weeks. Behaviour observations (not reported here) showed SFP progressed to IP, and in some pens hen death from cannibalism or permanent removal of pecked hens were recorded. IP occurred especially around the rump/uropygial gland area. In two pens, regular wounding/cannibalism occurred and both pens were removed from the study on ethical grounds in week 30 and the remaining hens rehomed. Overall, by 41 weeks flock size had been reduced by over 20%.

There were no differences due to main effects or interactions on the number of hens with PD, or the number of PD sites per hen. In week 41 there was a weak difference (P = 0.087) due to the Light main effect on feather-score, with the Low light treatment having ‘worse’ PD than Normal treatment (back-transformed means {btm} of max. feather-score: 1.8 vs 0.9, respectively). There was no effect of TRM, and no Light × TRM interactions. Figure 2 shows the proportion of hens at 41 weeks with different levels of severity of PD at the 7 body areas, for the two main effects. However, at 41 weeks there was an effect (P = 0.017) of side of the shed on PD, with hens in pens with south- compared to north-aspect having worse max. feather-score (btm = 2.1 vs 0.7, respectively). Similarly, side of the shed affected the number of PD areas per hen (in weeks 24, 29, 34 and 41; P < 0.01), with hens in south-aspect pens having more PD areas than north-aspect pens (e.g., btm for week 41 = 2.0 vs 0.4 areas, respectively).

Analysis of hen mortality plus removals identified a weak effect of the Light main effect (P = 0.062), with a tendency for a higher loss of hens reared in the Low than Normal lux (22.7 vs. 16.7% of hens, respectively), and no difference due to the TRM main effect. However, there was a strong (P < 0.001) effect of side of the shed on mortality plus removals, with losses of 33.7% of hens from south- compared to 5.8% from north-aspect pens to week 41.

An important finding from this study was that, when an SFP outbreak occurred in the experimental flock, differences due to the Light and TRM main effects were relatively minor compared to the differences between pens on the different sides of the shed. Hence our hypotheses were not proven. Clearly, other (unknown) factors had greater influence on the initiation of SFP than the imposed rearing treatments. Resultant PD, which was mainly evident on the rump, back and tail areas of hens, appeared to predispose hens to IP and cannibalism. Flock size consequently declined by ~20% of hens between weeks 16 and 41, with most dead or wounded hens receiving pecking injuries to the uropygial glands. In attempting to identify factors associated with the SFP/IP outbreak, there was a major rain event (133.6 mL) over 12 days during winter (late June to early July) that preceded the onset. Further investigation to identify differences in stimuli (and associated stressors) present between pens on the (colder/damper) south compared to (warmer/drier) north sides of the shed, could inform future research to identify the underlying cause(s) of SFP.

In general, the imposed experimental rearing treatments did not affect hen growth. While some differences in body weight were found (e.g., at 14 weeks due to Light; at 18 weeks due to TRM), these were relatively transient and by the next weigh-day; differences were nonsignificant suggesting hens had adapted to the environmental challenges which influenced body weight. In contrast, the findings regarding differences in SFP/PD and IP/cannibalism due to side of the shed were unexpected. However, the findings could help contribute to our developing knowledge of this very complex behavioural and welfare problem.

ACKNOWLEDGEMENTS: The experiment was funded by the Australian Egg Corporation Ltd. We thank Jadranka Velnic, Laura Sutton and Mary Anne Cronin who cared for the birds, and Dr Kate Hartcher who helped with some of the experimental measurements.