L.E. EDWARDS1
P.H. HEMSWORTH2
G.J. COLEMAN3
1Animal Welfare Science Centre, University of Melbourne, Parkville, Victoria, Australia
2Animal Welfare Science Centre, University of Melbourne, Parkville, Victoria, Australia
3Department of Psychology, Monash University, Clayton, Victoria, Australia
Human-animal interactions were studied at nine Australian farms and nine US farms. Measures of hen fear of humans, human activity in the laying shed, productivity records and shed parameters were collected at each shed.
Significant differences between countries were found for shed parameters, human behaviours and bird behaviours, and after adjusting for country differences, significant correlations were found between some of these human and animal variables. While the sample size is small and caution is required in interpreting the results, these relationships support the hypothesis that a sequential relationship exists between human behaviour, bird behaviour and productivity, and that the opportunity exists to improve hen welfare and productivity by manipulating this relationship.
Introduction
Extensive studies in the livestock industries have shown marked between-farm variation in the fear responses of farm animals, including poultry, to humans (Hemsworth and Coleman, 1998). Furthermore, Barnett et al. (1992) found a negative inter-farm correlation between fear of humans and the productivity of the laying hens. Such negative correlations indicate that high levels of fear of humans may be an important factor limiting the productivity and welfare of livestock.
Studies in the dairy and pig industries have shown significant sequential relationships between the stockperson's attitudes and behaviour towards animals and the fear responses of animals toward humans (see Hemsworth and Coleman, 1998). The existence of sequential relationships between human and animal variables in the livestock industries indicates that the opportunity exists to modify stockperson attitudes and behaviour in order to improve livestock welfare and productivity, and such opportunities may also exist in the poultry industries.
Since the work by Barnett et al. in 1992, new strains of laying hens have been introduced to the Australian industry, and the use of fully-enclosed environmentally controlled sheds has increased. Due to the observed behavioural differences in hen strains (Fraisse and Cockrem, 2006), and the more intensely controlled environment of the fully enclosed laying sheds, the present impact of human-bird relationships is unknown. The aim of the present study was to examine human-bird relationships with the current strains of laying hen in the modern commercial environment.
Materials and methods
Between April 2005 and May 2006, field work was conducted at ten Victorian farms, two NSW farms and 9 US farms, providing a total of 29 study sheds. The Australian farms used the Hyline Brown and ISA Brown strains of hen, with one farm using an Ingham Hisex strain. The US farms used the Hyline White 36 strain, with one farm using a Lohmann strain. All hens were aged between 26 and 106 weeks of age. The preliminary results for the relationships between human behaviour and hen fearfulness at eighteen of these sheds are discussed.
The amount of avoidance behaviour that an animal displays to humans is considered indicative of the level of fear of humans that the animal is experiencing (Hemsworth and Coleman, 1998). The level of fear of humans in laying hens in the present study was assessed by measuring the withdrawal response of birds to an unfamiliar human using several behavioural tests (adapted from Cransberg et al., 2000; Hemsworth et al., 1993). The primary test used was The Approaching Human Test, and was conducted at every 10th cage along the aisle resulting in a total of 1233 focal cages from the US and 598 focal cages from Australia. This discrepancy in cage numbers is due to the large variation between Australia and the US in shed size. The Australian sheds housed between 1300 to 29000 birds whilst the US sheds housed between 40500 and 183500 birds. During the Approaching Human Test the observer moved directly in front of the focal cage for 5 s and then stepped away for another 5 s. This procedure was repeated twice over a 20-s period. During each 5-s period, the experimenter counted the number of birds with their heads extended through the front of the cage, the maximum number of birds that moved their head into the front 5 cm of the cage, and made a point count of the number of heads still at the cage front at the end of each 5-s period.
This test resulted in a large number of variables, and a Principal Component Analysis was used to reduce the number of variables studied to a manageable number by identifying a small number of components or groups of variables that were statistically interrelated. The two resulting variables accounted for 81% of the variance in the avoidance response. These data were converted to scores and labelled as the variables ‘FORWARD SCORE’ and ‘HEADS OUT SCORE’. A high score for either was considered to indicate a low fear of humans. Shed averages of these scores were used in analysis.
Stockperson activity in the shed was observed over a second period, and the total amount of time that stock people spent in different areas of the shed and the total number of human interactions with birds (tactile, close visual and auditory contact) were recorded. From these observations, the following behaviour variables were calculated: presence in the aisle without cage approach (‘AISLE PRESENCE’, frequency/m of aisle length), close approach to birds (‘CLOSE APPROACH’, frequency/m of aisle length), hand in cages (‘HAND IN CAGE’, frequency/m of aisle length), total number of husbandry activities that occurred in the aisles (‘HUSBANDRY ACTIVITIES’, frequency/m of aisle length) and total time in shed (TIME IN SHED, seconds) . The duration of noise created by stock people using various equipment was recorded (‘NOISE’, frequency/m of aisle length). Frequency of the above behaviours was recorded using a five seconds bout criteria interval. The attitudes of the stock people toward the laying hens and their work were also assessed through the use of an attitudinal assessment questionnaire, however these data are not presented here.
Shed parameters such as aisle length, cage width, number of birds in the shed, space allowance and light levels were also collected. Productivity records were obtained where possible, however records for Australian farms, particularly for feed conversion efficiency, were not always available. The productivity variables included were peak hen day production (‘PHDP’), cumulative percent mortality at 29 weeks of age (‘MORTALITY’), peak feed conversion efficiency in kg of feed per kg of eggs (‘PEAK FCE’) and FCE at PHDP (‘FCEp’).
Results
Substantial variation in shed size and bird behaviour was observed between Australia and the USA.
An independent-samples t-test was conducted to compare the variables collected in each country (SPSS Version 14.0), and significant differences were found for shed parameters, human behaviours and bird behaviours. These country differences were adjusted for by subtracting the mean country value from individual variables for each farm prior to analysis.
Correlation analyses were conducted using the Pearson correlation. There were no significant relationships between shed parameters and any of the bird behaviour (fear of humans) variables and few significant relationships between bird behaviour and productivity variables. The correlations between the main human behaviour and bird behaviour variables are presented in Table 1, and the correlations between the main human behaviour and shed variables are presented in Table 2.
Discussion
While the sample size is limited, there are some consistent findings that support the hypothesis that human interactions may affect bird welfare and productivity.
Fear is generally considered an undesirable emotional state of suffering in both humans and animals (Jones and Waddington, 1992) and one of the key recommendations proposed to the United Kingdom Parliament by the Brambell Committee in 1965 (Anonymous, 1965) was that intensively-housed livestock should be free from fear. In this study a number of human behaviours were associated with fear. The duration of noise in the sheds was correlated with bird fear, with increasing duration of noise associated with increased fear. The majority of noise observed in the sheds was related to the use of air hoses and leaf blowers used for cleaning and indeed Campo et al. (2005) found that 1 hour of exposure to a 90dB noise stimulus resulted in an increased stress (heterophil to lymphocyte ratio) and fear (tonic immobility duration) response in laying hens. In contrast to the behavioural data, the duration of noise in the shed was positively correlated with peak HDP. Whilst loud noise may be aversive, a clean shed may improve HDP.
Increased frequency of some stockperson behaviours, such as presence in the aisle, close approach and hands in cages, was correlated with reduced fear. Placing a hand in the cage would have been considered by many to be an intuitively aversive procedure for the hens.
While few fear variables were correlated with productivity, a number of human variables corresponding to the amount of human contact, such as time in shed, presence in the aisle and husbandry activities, were related to productivity and mortality. Increased human contact was associated with better peak FCE and lower mortality. These correlations may reflect increased opportunity for stock people to identify and address production problems.
In contrast, increased presence in the aisle was associated with poorer FCE at peak HDP. Explanations for this correlation are not obvious.
While the number of farms studied is small, the observed relationships between human behaviour, bird behaviour and productivity provide limited support for the hypothesis of a sequential relationship between human behaviour, hen fearfulness and productivity in the egg industry. However, it should be noted that few fear variables were associated with productivity. This work provides a basis for further research on the effects of human behaviour on bird welfare and productivity, and methods for reducing fear of humans in laying hens. The human behaviour and fear relationships found in the Australian and US industries are an important finding since the need to reduce fear to improve welfare is becoming increasingly recognised.
Acknowledgements
Financial support of the Australian Poultry Cooperative Research Centre, the University of Melbourne and the Department of Primary Industries for this research is gratefully acknowledged.
References
Anonymous. (1965). Report to the Technical Committee to Enquire into the Welfare of Animals kept under Intensive Livestock Husbandry Systems. Committee Chairman Brambell, F.W.R., Her Majesty's Stationary Office, London, UK
Barnett, J.L., Hemsworth, P.H. and Newman, E.A. (1992). British Poultry Science, 33: 699710. Campo, J.L., Gil, M.G. and Davila, S.G. (2005). Applied Animal Behaviour Science, 91: 75 84.
Cransberg. P.H., Hemsworth, P.H. and Coleman, G.J. (2000). British Poultry Science, 41: 272-279.
Fraisse, F. and Cockrem, J.F. (2006). British Poultry Science, 47: 110-119 Hemsworth, P.H., Barnett, J.L. and Jones, R.B. (1993). Applied Animal Behaviour Science,
36: 197-210.
Hemsworth, P.H. and Coleman, G.J. (1998). Human-livestock Interactions: The Stockperson and the Productivity and Welfare of Intensively-farmed Animals. CAB International, Oxon UK.
Jones, R.B. and Waddington, D. (1992). Animal Behaviour, 43: 1021-1033
From Proceedings of the “19th Australian Poultry Science Symposium”, New South Wales, Australia.
