Pekin ducks are motivated to lay in their preferred nest substrate

Abstract

Introduction

Nest design is one of several factors contributing to floor-laying in farmed poultry, as nest features can determine if a laying bird finds the nest suitable (Appleby et al. 2004; Mench 2009). In chickens (Gallus gallus domesticus) and quail (Coturnix japonica), nest site selection can be influenced by the type of nest substrate available (Huber et al. 1985; Appleby & McRae 1986; Hughes 1993; Schmid & Wechsler 1998; Struelens et al. 2005; Guinebretière et al. 2012). In Pekin ducks (Anas platyrhynchos), the choice of nest can be influenced by enclosure and the presence of eggs (Makagon & Mench 2011; Makagon et al. 2011), but the importance of substrate to nest site selection has not been investigated.

Preference tests ask an animal to choose between different resource alternatives. Such tests have been used to establish preferences for nest substrate in a range of farmed animals. Chickens show preference for substrate that they can manipulate or peck at, such as peat or straw, rather than artificial substrates, such as plastic mesh or artificial turf (Huber et al. 1985; Struelens et al. 2005). Quail prefer nests containing hay over astroturf, while chaff tends to be more acceptable than hay (Schmid & Wechsler 1998). In fish, captive male Nile tilapia (Oreochromis niloticus) prefer sandy substrate over stones or no substrate for spawning (Mendonça et al. 2010). Rabbits (Oryctolagus cuniculus) prefer substrates such as straw or fine fibre over wood-shavings, the latter being commonly provided as nesting material on rabbit farms (Blumetto et al. 2010; Farkas et al. 2018). The conduct and interpretation of preference tests can be confounded by factors such as familiarity, learning ability, environmental cues, or other conditions that can influence an animal’s choice (Kirkden & Pajor 2006; Fraser & Nicol 2011). Notwithstanding these issues, preference tests can provide valuable insight into what an animal favours. Studies such as those noted above clearly demonstrate that an animal’s preference for substrate can be tested, and that the provision of resources in farming systems can be optimised if they are based on those preferences.

However, preference testing provides no insight into how important an animal considers a resource. Knowing how motivated an animal is to use a preferred resource provides insight into whether providing, or not providing, a preferred option will enhance or diminish animal welfare (Dawkins 1983). Behavioural demand tests measure how important an animal considers a resource by asking an animal to work increasingly hard to gain access to that resource (Mason et al. 1998). The combination of preference and motivational testing has been used to assess the motivation of hens for a preferred level of nest enclosure (Kruschwitz et al. 2008), a hen’s willingness to work for different foraging substrate (Gunnarsson et al. 2000), and the amount of work that pigs (Sus scrofa) are willing to perform for access to various rooting materials (Pedersen et al. 2005).

The provision of a suitable nest to farmed ducks should encourage positive welfare in laying birds, because nesting behaviour is highly internally regulated (Mench 2009). The decision to use a particular nest could be considered an active, goal-directed behaviour, and the fulfilment of such behaviour can contribute to a positive affective state. Failure to achieve the goal may result in negative states, such as frustration (Mellor 2015). To understand the response of an animal to its environment (and the subsequent interpretation of affect) multiple physiological and behavioural variables can be measured, preferably using a ‘hands-off’ approach to avoid inadvertent stress and the confounding of data (Cook et al. 2000). In birds, the measurement of corticosterone in the egg, instead of corticosterone in the plasma (the latter of which requires physical handling and blood sampling) has been used as a non-invasive method to assess birds’ response to acute stressors such as handling, ambient temperature increases, and movement of hens between cages (Downing & Bryden 2008). Correlation between the concentration of corticosterone in plasma and that in albumen has been demonstrated (Downing & Bryden 2008). The use of corticosterone in egg albumen as an indicator of stress has been trialled once in Pekin ducks that were denied access to a nest site (Barrett et al. 2021). That study found that any stress experienced was not reflected in the level of corticosterone in the albumen. As that was the first such study in Pekin ducks, further trials are warranted to better determine if the corticosterone concentration in egg albumen might be a useful physiological indicator of stress.

Another useful tool to assess the physiological effect of stressors on an animal is change in core body temperature to detect stress-induced hyperthermia (SIH). Animals exhibit SIH in response to many stressors, such as handling (Bittencourt et al. 2015), prolonged restraint (Gray et al. 2008), tests of fearfulness (Pedernera-Romano et al. 2010), and social stress tests (Kohlhause et al. 2011). Pekin ducks exhibit SIH when they are unable to access an established nest site (Barrett et al. 2021), a response considered to be related to frustration.

Frustration can occur if a strongly motivated behaviour is thwarted, resulting in a negative affective state for an animal. Changes in behavioural patterns can be used to gauge an animal’s subjective experience. In chickens, behaviours that have been investigated for their potential association with frustration include pacing (Wood-Gush 1972; Mills & Wood‐Gush 1985), preening (a displacement activity) (Duncan & Wood-Gush 1972; Mills & Wood‐Gush 1985; Meijsser & Hughes 1989), comfort behaviours such as head-shaking, tail-wagging or feather-raising/fluttering (Duncan & Wood-Gush 1972; Mills & Wood‐Gush 1985), feeding and drinking (also considered displacement behaviours) (Mills & Wood‐Gush 1985; Meijsser & Hughes 1989; Sherwin & Nicol 1993a), feather-pecking (Dixon et al. 2008), and redirected pecking behaviour (Kuhne et al. 2011). Wing-flapping has been associated with frustration in pigeons (Columbia livia) (Terrace 1972). As far as it can be determined, behavioural indicators of frustration have not been explored in the Pekin duck.

In the Australian duck industry, breeder ducks are typically provided with substrate such as sawdust or wood chips in the housing and in the nests. It has been estimated that floor eggs account for 20% of egg production in Australian duck farms (Luv-a-Duck, personal communication 2013). A better understanding of the factors that contribute to floor-laying in ducks could assist the industry to reduce the incidence of floor-laying and the associated costs. It is thus relevant to explore whether Pekin ducks have substrate preferences and how important that preference is. Thus, the use of preference and behavioural demand techniques together seems a logical approach to determine the significance of a preference for nest substrate in ducks.

With this study we aimed to:

• Determine if Pekin ducks exhibit nest substrate preferences;
• Assess ducks’ motivation to access their preferred substrate by asking them to perform an operant task (weighted push-door) to access their preferred substrate; and
• Assess the response to the stress of restricted access to the preferred substrate, by looking at changes in core body temperature, egg albumen corticosterone, and behaviour.
• It was hypothesised that:
• Ducks will show a preference for one type of nest substrate above others;
• Ducks will perform increasing amounts of work to access their preferred substrate; and
• Even if a less preferred substrate is available for free, ducks will show signs of stress, likely due to frustration, when they are unable to access their preferred substrate.

Materials and methods

Training and habituation in the behavioural demand apparatus

After arriving at the research facility, the birds had two weeks of habituation, as described in Barrett and Blache (2019). Handling was performed in a small temporary holding area for individual physical examination every second day during the first week, and then daily during the second week. Each pen had one empty behavioural demand unit (BDU) (Figure 1). During habituation, the end wall was removed from the unit, there were no nest-boxes in place, and the doors were open, so that the ducks could explore the unit unencumbered from the time of their arrival.

Figure 1.

Showing diagram (top) and photograph (bottom) of the behavioural demand unit that was used to test the substrate preference of laying Pekin ducks, and their motivation to access a preferred substrate.

All 15 ducks underwent training to learn how to use the door of the BDU. For three consecutive days, on three times each day, each bird was held in one part of the unit and provided the opportunity to exit through the open BDU to return to her social group. On the subsequent three days, the process was repeated, but this time the birds were required to pass through the closed, unweighted door. Finally, for the last three days, the birds were required to pass through a weighted door (20% of the group average bodyweight) to return to their social group. Thus, each bird had 27 opportunities to learn to exit from the BDU. At all stages of the training, if the bird had not exited the BDU within 2 min, she was returned to the group until the next attempt. The 12 birds that had learned most effectively how to the use push-door were taken through the remainder of the experiment. The selection criterium for this decision was the 12 birds with the highest proportion of successful exits from the box when the door was closed (nine when the door was unweighted, and nine when the door was weighted with 20% weight, for a total = 18 opportunities to exit; see Results; BDU training). Social group was used as the training reward because it was easy to apply and has previously been demonstrated to be effective (Barrett & Blache 2019). No assumption was made about the relative intensity of motivation for the social group versus a preferred nest substrate.

Since the ducks were to be housed individually in a BDU overnight for preference and behavioural demand tests, they were progressively habituated to spend more time by themselves in a BDU. This habituation aimed to mitigate any stress due to social isolation that might occur and followed that outlined in Barrett et al. (2021). To briefly review, each bird was placed in the BDU for an increasing time over a period of six days, until they spent 3 h in the unit. Food and water were available in the unit during this time, and individuals were within sight and sound of their social group.

Results

Discussion

The aims of the study were to determine if Pekin ducks display a preference for nest substrate, to assess the importance of that preference by asking them to exert increasing work to access that preferred substrate, and to assess their responses when they were unable to access their preferred substrate. The hypothesis that ducks have nest substrate preferences was supported by the results of the preference test, and the pattern of nest use and interaction when the birds had open access to a choice of substrates. The hypothesis that ducks are willing to work for their preference, and will show signs of stress when they are unable to access their preferred nest, was supported. The results of the behavioural demand test indicate that half of the ducks (6 of 12) were highly motivated to access the nest that contained their preferred substrate. The ducks that were willing to push 120% of bodyweight to access the nest experienced SIH and changed their behaviour when they were unable to access their preferred nest. However, any stress that occurred when they were unable to use their preferred nesting substrate was not reflected in changes to the concentration of corticosterone in egg albumen.

The overall rank order of substrate preferences was sawdust > hemp fibres > astroturf. Further evidence of substrate preference was found in the greater proportion of time that the birds spent in the nest-box that contained the preferred substrate (sawdust), or in the immediate area around this box, and the greater number of nest investigations and visits directed to this nest. One interpretation of these results is that, when offered a choice, Pekin ducks preferred to use a nest that contained substrate that could be manipulated. Similar preference for manipulatable nest substrate has been shown in chickens (Huber et al. 1985; Rietveld-Piepers 1987; Appleby et al. 1988). Farrowing sows also prefer nesting areas that contain straw over those that have structural elements of pre-formed nests (e.g. a hollow) but do not contain substrate (Arey et al. 1992).

In chickens, it is thought that fewer nest investigations of longer visit duration are indicative of preference, whereas higher numbers of nest investigations may indicate doubt (e.g. Meijsser & Hughes 1989; Struelens et al. 2008). As this is the first study of nest substrate preference in ducks, it is not yet clear if our findings are typical for this species. However, we note that once a duck passed through the closed door, she spent 90% of the observation period in the preferred nest regardless of door workload. This suggests that ducks were quite settled once in the nest-box. The low use of the astroturf nest when access to the sawdust nest was blocked would also support the notion that the low number of investigations and visits were indicative of the aversiveness of astroturf.

The results from the preference testing indicate that of the two manipulatable substrates, the ducks generally preferred sawdust over hemp fibres, though variations between individuals existed. The reasons for the variation are not entirely clear, but may include differences in particle size (hemp fibres being longer and larger), scent, or perhaps familiarity due to previous exposure to sawdust in the early stages of on-farm rearing. The issue of familiarity in preference testing is well recognised: animals may be attracted to, or avoid, environments that they have had previous experience with, though this may change over time as they become familiar with other available options (Fraser & Nicol 2011). It should also be noted that the presence of aspen chips in the nest-boxes, early on in the experimental period, could have contributed to the formation of a preference in the ducks, being somewhat similar to sawdust in particle size. Whilst it is difficult to know whether previous experience with sawdust or aspen chips influenced the initial choices made by ducks in the preference test, the repeated exposure to all substrates, and individual variation for choice of sawdust over hemp, suggest that familiarity was not the key driver in ducks choosing sawdust overall. A larger differential between sawdust and hemp, as well as sawdust and astroturf, would have been expected if previous experience with sawdust had an influence on a duck’s preference.

Most of the ducks were motivated to use a preferred nest substrate. Six of the 12 birds completed all workloads and attempted to pass the door to the preferred substrate even when it was blocked. Another four birds were willing to overcome workloads of between 40–100% bodyweight for access to the preferred substrate. For those ducks that completed all workloads, SIH was first recorded when they were required to push 100% of their bodyweight, a result similar to a previous study (Barrett et al. 2021), where 80% was the first workload at which SIH was observed. It is possible that the SIH seen at 100, 120% and the first night of the blocked door was due in part to the heat production that is associated with physical exertion. However, physical exertion cannot explain the results on nights 2–4 of the blocked door, because the birds still developed SIH despite making fewer attempts at the blocked door, and more at the unweighted least preferred substrate door, which would have required much less effort. Thus, the results indicate that if a bird is highly motivated to use a preferred nest substrate, denying the bird access to that resource elicits a stress response. This is supported by the lack of difference in the AUC between the first night of the blocked door and the following nights 2–4. The two birds that stopped pushing at 80 and 120% also showed SIH on nights when they made no attempts at the door. This finding suggests that they remained motivated for the nest but had surpassed their physical ability to access it, but consequently experienced stress, most likely due to frustration, when unable to access the preferred substrate. Although the small sample size in this study places limitations on how widely these results can be extrapolated to farmed duck populations, these findings, in conjunction with those of Barrett et al. (2021) provide preliminary evidence that further exploration of SIH could be a useful indicator of frustration in the Pekin duck.

The behavioural observations that we made here can provide a baseline of information from which behavioural indicators of frustration in Pekin ducks can be further explored. A majority of ducks showed an increase in the number of attempts and looks through the door to the preferred substrate, and an increase in the rate of wing-flapping and BDU pecking on the first night of the blocked door. Increased rates of wing-flapping and BDU pecking were seen in several individuals across all nights of the blocked door. While these results should be interpreted with caution, owing to small sample size and the limited ability for a robust statistical analysis, they do provide information on changes in behaviour that might be useful indicators of frustration in ducks. These behaviours have been shown to be indicators of frustration in other bird species. In pigeons, wing-flapping is increased during the early stages of a discrimination task in birds that have not yet learnt to correctly predict the presence or absence of a reward (Terrace 1972). Re-directed pecking in hens increases under operant learning conditions, when the hens are required to learn reversal of a task, or the task is not rewarded at all (Kuhne et al. 2011). It is also interesting to note that food-deprived hens show increased aggression towards a conspecific when they can see, but not access, food. This aggression takes the form of pecking and gripping the other bird with its beak (Duncan & Wood-Gush 1971). It is worth considering that the ducks’ pecking behaviour in the BDU may have been a form of redirected pecking. There is substantial evidence to indicate that not rewarding a previously rewarded task leads to a state of frustration, and subsequent aggression between birds (Papini et al. 2019). If such behaviour is considered in a commercial farming context, it can be theorised that where competition for resources exists (food, preferred nests, mates), or the environment does not allow the expression of behaviours that animals are highly motivated to perform, then aggressive interactions or abnormal behaviours, due to frustration, can be expected. Competition for nests has previously been proposed as a contributing factor to floor-laying in Pekin ducks, with 57% of nest exits being associated with an aggressive encounter (Barrett et al. 2019). Further studies are warranted to explore if redirected pecking occurs in ducks subjected to other frustrating situations, and whether this manifests differently in individual vs social contexts.

The limited change in pacing in potentially ‘frustrated’ ducks contrasts with previous findings in hens, where increased pacing was thought to be an indicator of frustrated nesting behaviour (Wood-Gush 1972; Sherwin & Nicol 1993b; Yue & Duncan 2003; Cronin et al. 2012; Tahamtani et al. 2018). A possible reason for the limited change in pacing seen in our ducks may be the different method of data collection. The present study recorded the rate at which pacing occurred, whereas those other studies counted the number of steps taken by birds. Thus, it would be useful in future to assess the step count of ducks in potentially frustrating situations to better align the interpretation of data with that in other species.

Displacement preening has been associated with frustration in hens (Duncan & Wood-Gush 1972). In our ducks there was no change in preening behaviour at any stage of the experiment. Preening may not be a useful indicator of frustration in ducks, or frustration may not have occurred. It is possible that due to the small sample size, and the generally low frequency at which preening occurred, an accurate representation of displacement preening was not gained in our experiment. The typical diurnal pattern of preening in Pekin ducks has yet to be established. Thus, preening as a possible indicator of frustration in ducks requires further exploration.

While we acknowledge the limitations of sample size, it is interesting to note the association between nest use and SIH when the ducks were exposed to the blocked door. Birds 2 and 4 chose to enter the least preferred substrate nest on either night 1 (bird 4) or 2 (bird 2) of the blocked door, and all subsequent nights thereafter. The SIH decreased in those two birds after night 2. Bird 7 exhibited SIH on all four nights of the blocked door, but chose to use the least preferred astroturf on the last night, while birds 6 and 11 exhibited SIH on all nights of the blocked door, and never used the alternative nest. There are two possible explanations for these differences. The fact that birds 6 and 11 chose never to use the least preferred substrate might be because their failure to pass the blocked door was such a negative experience that it impacted their willingness to try the alternative door. Similarly, in a previous study, ducks made no further attempts to pass a blocked door after failing at it on the first night (Barrett et al. 2021). An alternative explanation is that birds 2, 4, and 7 elected to use the least preferred substrate on the basis that it was the most adequate option available to them. These birds showed SIH on the first night that the door to sawdust was blocked and used the astroturf, suggesting that on that occasion they were frustrated by not being able to access the preferred substrate. However, the lack of SIH in birds 2 and 4 on subsequent nights suggests that, to them, there was no longer any frustration and that astroturf was better than laying on the floor. Thus, it may be that the astroturf was considered the ‘best of the worst’ nest substrate on offer. In contrast, birds 6 and 11 may have viewed the astroturf as equally, or more undesirable, than the floor, and never passed the alternative door. The diminished SIH in birds 2 and 4 could be explained by the process of habituation, where an animal’s response to a stressor decreases with repeated exposure. If habituation did occur, this suggests that birds 2 and 4 showed greater resilience about their choice of nest substrate than did birds 6 and 11. Possible explanations for an individual’s resilience may include genetic predisposition, broader early life experience, or variations in individual personality traits.

Although some ducks used the astroturf nest on occasion, the overall acceptability of this substrate was low. Besides being the lowest ranked substrate in the preference test, six of the 12 birds never used the astroturf nest during the behavioural demand tests. From a management perspective, astroturf is an attractive choice in poultry farming, because it is easier to provide and maintain than loose litter material. However, when making decisions about the type of resource to provide for animals, a straightforward consideration of whether animals will use that resource may be insufficient to meet welfare needs. Instead, the aim should be to provide resources that encourage a positive affective state by allowing the full expression of behaviours that an animal is motivated to perform. The expression of nesting behaviour is affected by the type of substrate that is provided to an animal. For example, chickens express nesting behaviour most fully when they are provided peat, compared to astroturf or wire mesh, with wire mesh resulting in a restless pattern of nesting behaviour (Struelens et al. 2008). In farrowing sows, the provision of straw allows for the most complex expression of nest-building and reduces stereotypic behaviour, compared to peat or no substrate at all (Rosvold et al. 2018).

In Australian duck production systems, it is typical for the barn floor and nest-boxes to be lined with some form of manipulatable substrate, with nesting bowls within the boxes (L Barrett, personal observation 2013). However, automated duck nests that are available internationally use either astroturf nest pads (Potters Poultry, Rugby, UK) or soft rubber mats (Vencomatic Group, Meerheide, The Netherlands). Given the strong preference for manipulatable substrates that was shown by ducks in the present study, the commercial availability of non-manipulatable substrates in large automated systems, and evidence in other species of the impact that substrate has on the expression of nesting behaviour, future studies should investigate whether the expression of duck nesting behaviour is influenced by the type of nest substrate that is available.

The analysis of corticosterone concentration in egg albumen was consistent with the results of Barrett et al. (2021), where no differences were found due to either increasing workload or psychogenic stress. Based on the findings of the two studies together, it is concluded that although changes in body temperature indicated that a stress response was more likely to occur with increased physical exercise and the inability to access a preferred nest site, the systemic corticosterone did not reflect that stress. It is possible that any change in corticosterone was not large enough and/or long enough to be reflected in the concentration of corticosterone in the albumen.

Conclusion

Laying Pekin ducks exhibit a preference for manipulatable nest substrate. Astroturf was found to be generally undesirable to the birds for the purpose of nesting. Most ducks were willing to expend increasing effort to pass through a push-door to gain access to the preferred nesting substrate, and they exhibited SIH when they were unable to do so. Changes to behavioural indicators suggest that the ducks experienced some frustration and provide an initial reference point for future studies of frustration-related behaviours in this species. We acknowledge that the small sample size makes statistical analysis challenging for some parameters. Similar to previous findings, the measurement of egg albumen corticosterone was not a useful indicator of stress in this situation.

The results indicate that in a commercial farming context a manipulatable substrate should be provided to nesting ducks. Practical considerations, such as the ongoing availability, hygiene, and cost of substrate are all valid issues that need to be considered. As this was the first report of ducks showing nest substrate preferences, further work is required to validate these findings and explore how attractive ducks find different substrates in commercial settings, and whether their preferences can help to mitigate floor-laying in commercially farmed flocks.

Acknowledgements

References