Estimation of optimal methionine, glycine, and tryptophan levels to improve plumage condition in ISA Brown laying hens

Supplementation of amino acids to reduce feather pecking behaviour and improve plumage condition has been previously studied and found to be successful (Savory, 1998, van Hierden et al., 2004). Feather eating behaviour performed by feather pecking birds suggests that feathers (predominately composed of keratin protein) may be consumed as an alternative source of amino acids, if digested.

Methionine is an essential amino acid required for egg production and feather development/growth. Although feathers contain low amounts of methionine, it may be required for feather synthesis and regeneration after feather pecking damage. Glycine is a non-essential amino acid and has a substantial presence in feathers, but requires dietary inclusion due to a high excretion rate and insufficient in vivo synthesis (Kleyn, 2013). Added Tryptophan has been reported to reduce gentle (van Hierden et al., 2004) and severe (Savory, 1998) feather pecking behaviour; however its use in combination with additional amino acids is unknown.

The aim of this trial was to determine an optimal combination of dMet, dGly, and dTrp levels that could effectively improve plumage condition. The trial used a Box Behnken experimental design which is based on a second order three-level incomplete factorial design and allows for the use of response surfaces and multivariate optimisation (Ferreira et al., 2007).

A total of 156 ISA Brown hens (68 weeks old) was used in this trial. Birds were selected from a previous trial based on observed feather pecking behaviour directed at artificially presented feathers. Birds were housed in group cages containing two feather pecking and two non-feather pecking hens and given one week to acclimatise to new housing. Each group cage was randomly allocated to one of 13 dietary treatments set according to a three-factor, three-level Box Behnken Design with three replicates per diet. The centre point levels of dMet, dGly, and dTrp inclusion as analysed were 0.49, 0.76, and 0.23%, supplemented levels were 0.56, 0.88, and 0.26% and reduced levels were 0.43, 0.65, and 0.20%, respectively. Requirement levels of dMet, dGly, and dTrp were 0.51, 0.59 and 0.21%, respectively. Feed was provided ad libitum to each group cage. Amino acid analysis of both the excreta and diets was conducted by Evonik AMINOLab, Singapore.

Feather pecking was measured directly via behavioural observations of hens in the group cages, and indirectly via plumage condition scoring at the beginning and completion of the trial. Each hen was scored using a feather scoring system adapted from Tauson et al. (2005). Plumage condition was assessed on six body regions (neck, breast, back, wings, tail, and vent) with each region assigned an ordinal score from 1-4 with “four” being perfect plumage condition and “one” completely denuded. Plumage condition of each region was perfect at the beginning of the trial in all regions except the neck. The neck region was excluded from analysis due to plumage damage caused by rubbing of the neck feathers on the cage during feeding rather than feather pecking. Feather score data were analysed using proportional odds linear regression MASS and RESHAPE2 packages in R (R Foundation for Statistical Computing, Austria). Feather pecking behaviour observations were conducted twice daily on each cage from days 14-28. Observations were conducted for two minutes per cage, and all instances of feather pecking behaviour and severity of pecking bout were recorded. Feather pecking bouts not directed towards the aforementioned body regions such as the comb, wattle, head, or feet were categorised as “other” and excluded from the analysis due to difficulty in condition assessment.

A total of 254 feather pecking incidents was recorded over the course of the trial. The proportion of ‘gentle’ and ‘severe’ feather pecking bouts, out of all feather-pecking bouts directed at each body region, is displayed below (Table 1). As shown in Table 1, about twothirds of bouts were classed as ‘gentle’, and overall about one-half of observed bouts were directed at the neck region of hens.

A proportional odds linear regression model was fitted to the sum of feather scores per bird at the end of the trial. Regression analyses indicated linear effects of dMet (P < 0.001) and dTrp (P = 0.03) inclusion but no quadratic, or interactive effects of dMet, dGly, or dTrp inclusion on the sum of feather scores. Using the RESHAPE2 package in R, the probability of a bird obtaining each overall feather score (ranging from 16-20) for every level of dMet, dGly, and dTrp could be predicted (Figure 1).

A generalised linear regression on the predicted probability of observing perfect plumage condition based on dMet, dGly, and dTrp was conducted. All predictors were found to be significant (P < 0.001, P < 0.001, and P < 0.001 respectively), with the probability of observing plumage damage decreasing as dMet, dGly, and dTrp inclusion rates increased. The minimum dMet, dGly, and dTrp inclusion rates required to ensure the majority of birds maintained perfect plumage condition were predicted at 0.46, 0.71, and 0.22%, respectively. The maximal likelihood of observing perfect plumage condition (61.31%) was predicted when dMet, dGly, and dTrp inclusion rates were 0.56, 0.88, and 0.26%, respectively.

Proportional odds linear regression analysis on the plumage condition of individual body regions indicated a weak effect (P = 0.06) of dGly inclusion on breast region feather score, and a strong effect of dTrp inclusion on back region feather score (P < 0.001). The predicted probability of observing each plumage condition score for the back region are displayed in Figure 2. The maximal likelihood of obtaining perfect breast (99.95%) and back (99.68%) plumage feather scores was predicted when dMet, dGly, and dTrp inclusion rates were 0.56, 0.88, and 0.26%, respectively.

The significant effect of methionine supplementation on overall plumage condition in this trial contradicts the findings of Kjaer & Sorensen (2002), who observed no improvement in integument condition when methionine + cysteine content was supplemented up to 0.82% inclusion.

To date, no other studies have reported on the effect of glycine inclusion levels on feather pecking behaviour. However, results of this trial suggest only a slight effect of dGly inclusion on plumage condition on the breast region.

Tryptophan supplementation of up to 2% has been previously found to reduce gentle (van Hierden et al., 2004) and severe (Savory, 1998) feather pecking behaviour. Since dTrp supplementation in this trial resulted in a significant reduction in back region plumage damage and increase in overall plumage condition, this study supports the notion that feather pecking behaviour is positively influenced by tryptophan inclusion.

The results of this study indicate that reduction of dMet, dGly, and dTrp inclusion levels below recommended requirement (0.51, 0.59 and 0.21%, respectively) tends to increase the likelihood of plumage damage occurring. Similarly, increasing inclusion rates above recommended requirements tended to the increase the likelihood of perfect plumage condition scores.

Supplementation above recommended dietary requirement of dMet, and dTrp significantly affected the overall plumage condition of birds across five body regions (breast, back, wings, tail, vent). The use of a Box Behnken design allowed for prediction of optimal levels of dMet, dGly, and dTrp for perfect plumage condition at 0.56, 0.88, and 0.26%. Moreover, the presence of plumage damage in the back region becomes more prevalent than no damage when dTrp levels are reduced below 0.20%.

Further study investigating the effect of amino acid supplementation in larger groups, with additional replicates, and over a longer time frame is recommended to overcome the limitations of this study which may have resulted in the low level of feather pecking observed.

ACKNOWLEDGEMENTS: We would like to acknowledge the contributions of Evonik Animal Nutrition, the Poultry CRC, and Poultry Research Foundation – University of Sydney for supporting this project.