Does the quantity of reward in a Y-Maze preference test affect the hen choice and motivation?

The quantity of reward in a Y-maze preference test refers to the period of time the animal is allowed contact with the chosen resource. Fifteen hens (Hy-Line Brown strain) were given eight preference test trials (conducted on alternate days) for their choice between dust (a tray of peat moss) or social contact (the presence of a familiar subordinate hen). The quantity of reward for social contact was 5 minutes for all birds. However, the hens were allocated to three treatments (n = 5), which differed in the quantity of reward when dust was chosen: Long (45 minutes), Intermediate (20 minutes) and Short (2 minutes). During testing, hens were housed individually and were deprived of both social contact and dust. Hens on all treatments chose dust significantly more often than expected at chance level (P < 0.001 for all). The choice behaviour was not significantly different between treatments, however the Intermediate hens had a tendency to choose social contact more often (P = 0.07). Intermediate birds also took significantly more time to move to their choice compared to both the Short and Long birds (P = 0.0015). In addition, there was a significant treatment effect on the proportion of dust-chosen trials in which a dustbathing bout was interrupted (P = 0.0079), with the Long treatment having the lowest proportion interrupted and the Short treatment having the highest proportion of dustbathing bouts interrupted. Results indicate that dust was an attractive resource in the Y-maze for all birds. The tendency to choose dust less often and take longer to move through the Y-maze suggests that Intermediate treatment birds were less motivated to gain access to dust. This could perhaps have been due to the period that these birds had access to dust, which was less than the average time for a dustbathing bout, causing a greater proportion of dustbathing bouts to be interrupted in comparison to the Long treatment. Therefore, while the quantity of reward did not significantly influence choice behaviour, the Intermediate treatment appeared to reduce hen motivation for accessing dust.

Animal preferences may tell us what is important to an animal and thus provide an indication of its welfare. Preference tests carried out in a Y-maze apparatus, where an animal makes a choice between two resources, may appear to be straightforward. However, aspects of the design of Y-maze preference tests may have the potential to influence motivation and thus choice behaviour, leading to spurious results that are not reflective of the animal’s true preferences.
One such factor is the quantity of reward. The quantity of reward in a Y-maze preference test refers to the period of time the animal has with its chosen resource. Although it has been previously proposed that the quantity of reward may affect how ‘attractive’ a resource is to an animal in a preference test (Nicol, 1997; Kirkden and Pajor, 2006) this concept has yet to be examined. This experiment set out to determine the effects of the quantity of reward on the choice behaviour and motivation of laying hens in a Y-maze preference test where hens were offered the choice between dust and social contact.
Dustbathing is a series of behavioural components that aids the maintenance of plumage (van Liere and Bokma, 1987; Olsson and Keeling, 2005). On average, the duration of a dustbathing bout is 27 minutes and occurs every second day with peak activity being around midday (Vestergaard, 1982). Laying hens have been found to have preferred substrates in which to dustbathe, with peat moss being the most preferred material (Petherick and Duncan, 1989; de Jong et al., 2007).
Laying hens are highly social animals that flock together and form a stable dominance hierarchy, which may remain stable for a number of years (Schjelderupp-Ebbe, 1922 cited in Mench and Keeling, 2001). Social contact is presumably important for laying hens as many behaviours such as feeding, dustbathing and preening are often performed in synchrony by many individuals (Hughes, 1971; Webster and Hurnik, 1994; Duncan et al., 1998). Additionally, social isolation is known to cause stress, both behaviourally and physiologically, in domestic chicks (Jones and Merry, 1988).
Thus a dustbathing substrate and social contact are important resources for laying hens. We hypothesised that the quantity of reward would impact on how attractive the resources were perceived by laying hens and this would ultimately impact on hen choice behaviour. To test this hypothesis, we gave a choice in a Y-maze for access to a familiar, subordinate hen for a fixed period of time and a dustbathing substrate for three different periods of time.

Fifteen hens (Hy-Line Brown laying strain) aged 31 weeks were housed in individual cages (0.57 m x 0.50 m x 0.48 m). Each cage contained a dustbath that was refilled daily with peat moss. Throughout the experiment, hens had ad libitum access to food and water. The birds were firstly familiarised to the Y-maze apparatus (for a description of the apparatus, refer to Laine et al., 2007). Hens were then trained on alternate days for a total of five training sessions. During training and testing, hens had their home cage dustbath removed and were deprived of social (visual) contact by placing opaque rubber partitions between cages. All training and testing sessions were conducted from 1100 h each day. Immediately following training, hens were tested for eight trials (one per day) conducted on alternate days for their choice between social contact (the presence of a familiar subordinate hen) or dust (a tray of peat moss). The quantity of reward for social contact was 5 minutes for all birds. However, birds were allocated to three treatments which differed in the quantity of reward for dust; Short (2 minutes), Intermediate (20 minutes) and Long (45 minutes). ‘Time to choice’ was defined as the time taken from when the start box gate was opened to when the hen moved into one of the Y-maze arms. A choice was defined as the resource that the Y-maze arm contained that the hen moved into on each test trial. A dustbathing bout was defined as interrupted if the hen was picked up and taken back to her home cage by the experimenter prior to the completion of the dustbathing bout.

Hen choice behaviour was compared to chance level (i.e. 50:50) using a Chi-Square test. After log transformation to reduce skewness of data, an analysis of variance (ANOVA) was used to compare the choice behaviour between treatments. The hens’ choice for dust was higher than would be expected by chance for all treatments (P < 0.001). However, the Intermediate treatment hens had a tendency (P = 0.07) to choose social contact more than the other treatments. The back transformed proportion of trials in which dust was chosen by the Short, Intermediate and Long treatments were 0.98, 0.83, and 0.98 respectively.
After negative reciprocal transformation, an ANOVA was used to compare the time to choice between treatments. The Intermediate treatment was significantly slower to make a choice compared to both the Short and Long treatments (P = 0.0015), with no difference between the Short and Long treatments (Fig 1).
After log transformation, an ANOVA was used to compare the proportion of dustchosen trials in which a dustbathing bout was interrupted between treatments. There was a significant treatment effect, with the Long treatment having the lowest proportion of dustbathing bouts interrupted and the Short treatment having the highest proportion of bouts interrupted (P = 0.0079). The back transformed proportion of dustbathing bouts that were interrupted for the Short, Intermediate and Long treatments were 1, 0.995 and 0.17 respectively.

The results indicate that dust was an attractive resource for all birds, as hens in all treatments chose dust significantly more often than would be expected by chance. Peat moss was chosen as the dust substrate as it has previously been found to be the preferred dustbathing substrate by laying hens (Petherick and Duncan, 1989; de Jong et al., 2007). In addition, preference test trials were conducted on alternate days, during the late morning/early afternoon to coincide with peak dustbathing activity (Vestergaard, 1982) and therefore presumed peak dustbathing motivation. These factors may have made very attractive, especially when hens were deprived of dust in their home cage. Moreover, even though the hens were deprived of social contact in their home cages, they had auditory and olfactory contact with other birds. Perhaps this contact, coupled with the brief visual contact hens had with the ‘social contact’ bird while they were in the start box of the Y-maze (prior to making a choice) was sufficient social contact for hens.
The speed of movement through the Y-maze and other tests where an animal must move to a resource can be equated to motivation (e.g. Petherick et al., 1992). It would be expected that an animal with a high motivation for a resource would move faster than an animal with low motivation. This, as well as having a tendency to choose dust less compared to the other treatments, suggests that birds in the Intermediate treatment were less motivated to gain access to dust in the Y-maze. This could be due to the quantity of reward that these hens had with dust (20 minutes), which is less than the reported average dustbathing bout a a b Aust. Poult. Sci. Symp. 2009...20 155 duration of 27 minutes (Vestergaard, 1982). In the present experiment, Intermediate treatment hens had a greater proportion of dustbathing bouts interrupted compared to the Long treatment. Mason et al. (1998) suggested that repeated interruption of an activity might devalue the resource. Perhaps repeated interruption of the Intermediate birds’ dustbathing bouts devalued the dust for these hens. If so, this devaluation could explain the apparent low motivation for dust access in the Y-maze.
Interestingly, the Short treatment did not differ from the Long treatment in terms of choice behaviour or time to choice. However, Short treatment hens had the highest proportion of dustbathing bouts interrupted compared to the other treatments. This repeated interruption, however, did not seemingly affect Short treatment hen choice behaviour or motivation (in terms of speed of movement) when compared to that of the Long treatment hens. Perhaps interruption at the very beginning of the dustbathing bout was less aversive to the hen compared to interruption further into the bout. This could explain why the small quantity of reward for dust seemingly had no impact on hens in the Short treatment.
While the quantity of reward did not significantly alter hen choice behaviour, it did apparently impact on motivation to access dust. Further research examining the effects of dustbathing bout interruption on motivation may aid in interpreting results obtained in this experiment. If interruption of an activity does prove to have an influence on animal choice behaviour or motivation, this may have important implications for the selection of reward quantities in animal preference tests in the future.

The authors gratefully acknowledge the financial support from the University of Melbourne, Australian Poultry CRC and the Victorian Department of Primary Industries.

De Jong IC, Wolthuis-Fillerup M, van Reenan, CG (2007) Applied Animal Behaviour Science 104, 24-36.
Duncan IJH, Widowski TM, Malleau AE, Lindberg AC, Petherick JC (1998) Behavioural Processes 43, 219-228.
Hughes BO (1971) British Poultry Science 12, 359-366.
Jones RB, Merry BJ (1988) Behavioural Processes 16, 75-86.
Kirkden RD, Pajor EA (2006) Applied Animal Behaviour Science 100, 29-47.
Laine S, Arnold NA, Hemsworth PH (2007) Proceedings, Australian Poultry Science Symposium 19, 28-31.
Mason G, McFarland D, Garner J (1998) Animal Behaviour 55, 1071-1075.
Mench J, Keeling LJ (2001) In: Social Behaviour in Farm Animals [LJ Keeling, HW Gonyou eds.] pp. 177-209. CAB International Publishing, Wallingford UK.
Nicol C (1997) In: Animal Choices [JM Forbes, TLJ Lawrence, RG Rodway, MA Varley eds.] pp. 35-44. British Society of Animal Science, Edinburgh UK.
Olsson IAS, Keeling LJ (2005) Applied Animal Behaviour Science 93, 259-282.
Petherick JC, Duncan IJH (1989) British Poultry Science 30, 229-238.
Petherick JC, Sutherland RH, Waddington D, Rutter SM (1992) Applied Animal Behaviour Science 33, 357-366.
van Liere DW, Bokma S (1987) Applied Animal Behaviour Science 18, 197-204.
Vestergaard K (1982) Applied Animal Ethology 8, 487-495.
Webster AB, Hurnik JF (1994) Applied Animal Behaviour Science 40, 153-165.