Understanding the behavioural needs of growing and finishing pigs and how we can meet those needs when designing future housing systems

Animal emotions – and what about the pig?

With some notable exceptions Darwin’s 1872 book on expressions of emotions in man and humans, ‘feelings’ or emotions have historically been viewed as non-scientific and not a subject fitting for scientific enquiry. However, during the last decades we have seen a resurrection of scientific interest in the field of mechanisms of emotion, not at least because of the increasing prevalence of emotional control deficits in human mental disorders. There are many definitions of emotions, most of them originating from human psychology. For example, Cabanac (2002) defines emotion as “any mental experience with high intensity and high hedonic content characterized as pleasure or displeasure”. He proposed that pleasure can be used to equate the strength of motivational drive that is linked to each regulatory need (Cabanac, 1992). Any behaving organism, being human or animal, must rank its priorities, and thus pleasure (or lack of suffering) can be a currency to achieve this ranking. Nevertheless, some authors still claim that consciousness is a prerequisite to experience emotions (e.g. Damasio et al., 2000), and there is still an ongoing debate concerning animals` capacity to be aware of their emotions (e.g. Salzen, 1998). Several authors within the field of evolutionary biology have hypothesised that emotions are processes evolved to help animals to avoid harmful stimuli and seek valuable resources or reward (e.g. Panksepp, 1994; Cardinal et al., 2002; reviewed by Paul et al., 2005). This view has been summarized by Balcombe (2009) in his paper on animal pleasure with: ‘evolutionary principles predict that animals, like humans, are motivated to seek rewards and not merely to avoid pain and suffering’. Today, in the human literature it is well described that emotions are essential for biosocial perceptions, serving to cast a selective attentional spotlight on some experiences (prioritizing versus distractions, incoherence versus focus, selective etc), that they are a motivator of behaviour. Positive emotions will enable an individual in decision making, i.e. predicting risk taking evaluations (Levine, 2007). In principle, pleasure can be categorized into four types (Duncker, 1940; reviewed by Cabanac, 2002).

1. Sensory enjoyment or displeasure i.e. to enjoy the stimulus or the consequences of behaviour;

2. Aesthetic enjoyment i.e. to strive for a better understanding;

3. Desire (e.g. for a steak, a book, a love, etc.), not a reaction but the fulfilment of a need;

4. Pleasure in achievement/problem solving, dynamic joy of succeeding, in victory.

At the simplest level, usually approach/avoidance can be used to measure valence (pleasant- or unpleasantness). Nevertheless, more complex behaviours such as play, and exploration are more adequate measures of positive affect because these behaviours usually only occur in the absence of fear and anxiety and after basic needs are fulfilled. Play behaviour of young animals as in humans stimulates motor and social skill development, which are basic components to secure a future positive, stable behavioural mental development and stimulates advanced coping abilities (reviewed by Estevez et al., 2007). The pleasure component of a play behaviour helps to motivate the animal/human to continue this behaviour like in explorative behaviour, solving a challenging task, enjoying something tasty or comfortable (i.e. tactile stimuli), or enjoying success of victory. For pigs, provision of straw, peat or other sources of rooting material may stimulate this behaviour as well as exploration, and for a longer period than simple play objects such as chains or tyres. Rooting behaviour is a highly prioritized behaviour in pigs. Although chopped straw is commonly used to stimulate this behaviour, peat, mushroom compost, sand, sawdust, wood shavings, branches, beets and silage are all ranked above straw in preference tests and operant conditioning tests (e.g. reviewed by Studnitz et al., 2007). The less manipulative objects like beams, ropes, tyres or chains are least preferred. This preference or ranking of different rooting materials may also be manifested through the different level of pleasure or indicators of positive affect that these elements create. Thus, studying behavioural indicators when presented different rooting materials, may serve as an efficient way of evaluating different sources of rooting materials as well as other sources of environmental enrichment. Since emotions are short-lived responses, spontaneous behavioural expressions in animals such as ear or tail movements, body postures or even facial- and eye expressions during play and exploration, are highly relevant to be carefully studied. Furthermore, the variety of vocalisations (in terms of intensity and category; e.g. Manteuffel et al., 2004) can be relevant indicators of emotions. Can behavioural indicators of affective states be viewed as animal signals? Well, what is the point of showing joy, satisfaction or happiness if you cannot share it with someone? And what is the point of showing distress or misery if you were all alone and did not think that there was someone around that could take care of you or help you out of this difficult situation?

“Fulfilment of a need” and the “success in solving a learning task” form the basis for initiating a positive emotional state (i.e. positive mental balance). Cognitive challenges presented in connection with foraging behaviours initiate positive emotional states in the animals and thus improve welfare (Puppe et al., 2007). An increase in environmental complexity that creates novelty and challenges in terms of increased movement, problem solving, or social stimuli also have positive effects on brain function and cognitive skills in general (Radak et al., 2001; Würbel, 2001). According to Manteuffel et al. (2009), ‘instrumental behaviour includes motivation for a specific reward, anticipation of its successful acquisition and positive appraisal’. Siegford et al. (2008) demonstrated that when pigs were successful in solving cognitive challenges during early life they showed reduced fear responses as adults, suggesting that positive experiences reduces negative emotions. Spatial learning trials in a maze have been successfully used in many species including and pigs (e.g. reviewed by Murphy and Arkins, 2007; Siegford et al., 2008) and pigs appear to solve these challenges rather well.

Emotions activating the limbic system has an impact on modulated immune response (Haas and Shauenstein, 1997). As greater negative affect in humans is linked to heart disease, cancer, arthritis and diabetes, greater positive affect is linked to lower hospital re-admission, reduced risk of stroke and a lower mortality (review: Consedine and Moskowitz, 2007). Humor and laughter also have a positive impact on recovery of cancer patients (Mahony et al., 2002). Although, health effects are documented in humans, their underlying biological mechanism has been scarcely studied in animals (Ernst et al., 2006). In principle, the effects of positive and negative psychological experience may depend on the degree of success or frustration.

The importance of environmental enrichment

Environmental enrichment can be defined as species-relevant modifications of animal environments resulting in improved biological functioning (modified from Newberry, 1995). Farm animals have strong behavioural needs that are important to satisfy in order to achieve a positive mental balance and thereby more healthy animals. Pigs are no exception from the rule. The need to play and explore by rooting in social groups of young pigs, are both highly prioritized activities and extremely important for their behavioural development into robust individuals. For instance, serum content of brain derived neurotrophic factor, BDNF, that plays an important role in neural survival, growth and plasticity, increases when young pigs are provided with foraging substrate block as enrichment during lactation or after weaning compared to pigs housed in barren environments (Rault et al., 2018). Greater concentration of BDNF in the brain is associated with an improved cognitive function and a greater stress resilience and could thus be an important welfare indicator.

Provision of straw, peat or other sources of rooting material may stimulate play behaviour as well as exploration, and for a longer period than simple play objects such as chains or tyres. In our recent project (Ocepek et al., under review), we offered weaned pig litters 10 l of rooting material (silage, long-stemmed straw, peat, or a combination of all three), twice daily, and compared this with groups that were not offered any rooting material. Behaviours considered indicative of positive affective states in this context (exploration, play, tail curled, tail wagging), as well as behaviours associated with harm (ear/tail manipulation, aggression, tail down), were quantified from video recordings. The peat and combo conditions resulted in higher levels of exploration, play, tail curled and tail wagging, and lower levels of ear/tail manipulation, aggression and tail down, compared to control, with the silage and straw conditions mainly giving intermediate results. Pigs showed more exploration, tail curled and wagging after than before provision of silage, straw, peat and combo, whereas an increase in play after material provision occurred only in the peat and combo conditions. Similarly, Marcet-Rui et al. (2019) found that straw was efficient in reducing harmful behaviours, but did not affect indicators of positive affective states, such as tail movement. In our study (Ocepek et al., under review), exploration occurred at similar levels on Day 1 and 4 of exposure to the peat and combo conditions whereas it declined across days in the other conditions. Comparatively, ear/tail manipulation and aggression increased in the silage condition. Peat as a single material and even more a combination of three rooting materials, substantially reduced negative behaviours and increased the occurrence of positive behaviours indicative of a more positive affective state in these pigs. The combo treatment also led the pigs to contact a novel person and a novel object faster, and enhanced the ability to collaborate with a pen mate in a small task to access rooting material from a closed box with a lid (Woldsnes et al., 2019). Therefore, we can expect stronger, positive and more long term effects on welfare if we provide pigs with a combination of materials at least twice a day. The amount of material should be sufficient, and for instance smaller amounts of 100 grams or less with chopped straw, tested experimentally, do not give clear welfare effects (Amdi et al., 2015). In Norway, it is manifested in the regulations that all pens need a small amount of sawdust to keep pens clean, but this is not considered as rooting material. The less manipulative objects like beams, ropes, tyres or chains are least preferred. Wooden sticks and especially spruce are more attractive than chains and other objects, but still not successful in reducing harmful behaviours such as ear chewing or tail biting (Chou et al., 2018). The preference or ranking of different rooting materials may also be manifested through the different level of pleasure or indicators of positive affect that these elements create. Thus, studying behavioural indicators when presenting different rooting materials, may serve as an efficient way of evaluating different sources of rooting materials as well as other sources of environmental enrichment.

Although only marginal physical (i.e. addition of toys, extra feeder) and social (i.e. larger groups) enrichment can make a temporary difference for pig welfare in terms of less body lesions caused by aggression, more exploratory behaviour, and an increased ease of handling (Tönepöhl et al., 2012), the goal for future housing systems should be to create more complex, heterogenous environments, including access to an outdoor areas (Figure 1), that can stimulate the pigs to a much larger extent than what we see in indoor systems today. It is not necessarily a discussion about indoor vs. outdoor systems since there are also many welfare challenges with an outdoor system, and theoretically it is possible to make a fully enriched environment indoors. However, the cost of housing pigs in larger groups with substantially more space per pig, with more equipment (although larger groups require fewer pen divisions) and material to stimulate activity, will cost more indoors than outdoors. We therefore predict that future housing for pigs will include access to controlled, outdoor areas that are protected from wild boars and pathogens.

Figure 1. Future housing of fattening pigs with access to a controlled outdoor arena with deep litter/straw bedding. The illustration is made by May Helen Gryte, Fjøssystemer A/S, Norway.

In an indoor environment, the pens are often barren and with too little stimulation between the time of feeding/drinking and resting. It appears that nothing can compete with access to straw bedding or other types of deep bedding areas as the most enriched environment for pigs (e.g. Scott et al., 2006), and should preferably be included in future housing systems to a larger extent if access to outdoors is not an option. For pigs to be able to fully express their behavioural repertoire, and especially play and explore, enough space is needed. Thus, we expect positive effects of environmental enrichment to become more evident when enough space is available. More specifically, growing pigs with more space available (1.0 m²/pig) explores rooting material (i.e. chopped straw or maize silage) more and manipulate pen fittings less than pigs with little space (0.64 m²/pig; Jensen et al., 2010). There is also a mounting body of evidence that suggests regular exercise improves brain function and causes structural, biochemical, and physiological adaptations via different pathways (Radak et al., 2007).

One of the most convincing studies regarding the effects of enrichment on production performance and meat quality was done by Beattie et al. (2000). In their study, pigs were reared from birth to slaughter in either barren or enriched environments. The barren environments were defined as intensive housing (slatted floors and minimum recommended space allowances) and the enriched environments contained extra space, and an area with peat and straw provided from a rack (i.e. not only one source of enrichment). Not surprisingly, this more complex, enriched environment reduced time spent inactive and time spent involved in harmful social and aggressive behaviour while increasing the time spent in exploratory behaviour. During the finishing period, mean daily food intakes were higher and food conversion ratios were lower for pigs in enriched environments. Growth rates were also higher for pigs in enriched environments during this period which resulted in a higher carcass weight. Pigs from enriched environments also had greater levels of backfat than those from barren environments. Finally, meat from pigs reared in barren environments was less tender and had greater cooking losses than pork from pigs reared in enriched environments. Whether pigs are housed in partially slatted or straw bedded systems does not seem to affect meat quality variables in other studies that much even though the deep bedding systems usually increase activity and locomotion (reviewed by van de Weerd & Day, 2009), but many social factors in interaction with type and amount of environmental enrichment could potentially affect meat quality variables. This must be studied in a more detailed and systematic way including the entire housing design making more heterogenous environments where the pigs can be occupied with different positive activities, but then again, we should not expect that all improvements in welfare are translated into improved production performance.

Social dynamics and the effects of animal density and group size

Pigs prefer to be kept in their stable litter groups as long as possible and mixing within a closed space is almost with no exception associated with more aggression and body lesions caused by fighting both in young pigs (e.g. Andersen et al., 2004) and in adult sows (e.g. Andersen et al., 1999). Therefore, the more stable we can keep the groups and the less mixing we must do, irrespective of animal density or group size, the better it is for pig welfare. Behavioural plasticity allows animals to change strategies and adapt more easily to varying environmental (social and physical) conditions within a confined group (Estevez et al., 2007). We developed a theoretical model that describes how aggression among unacquainted, weaned pigs is a function of group size when keeping space per animal constant (Andersen et al., 2004). We proposed that as the number of potential competitors increases, more individuals will not benefit from getting involved in costly fights as the probability of monopolizing resources will decline with increasing group size. Small group sizes of 6-12 individuals per pen have significantly more fights per pig than groups of 24 pigs, and in the largest group size a higher proportion of individuals did not engage in aggressive conflicts at all (Andersen et al., 2004). Other studies show similar results (Nielsen et al., 1995; Turner et al., 2001) or that the aggression remains relatively constant with increasing group size (Schmolke et al., 2004). Overall, it appears that pigs not only are less aggressive in larger groups, but also shift to a low-aggressive social strategy when they are moved from a small (18 pigs) to a large group (108 pigs), and that pigs raised in smaller groups tend to be more aggressive when meeting unacquainted pigs in a new group situation (Samarakone & Gonyou, 2009). In the literature we find terms such as ‘futures contracts for nonaggression’ (Pagel and Dawkins, 1997) and the ‘tolerant system’ (Estevez et al., 1997) to describe this phenomenon of reduced aggression in larger groups. Pigs kept in groups of 20 compared to 5, 10, or 15, also made fewer but longer visits to the feeder and ate more per visit and faster than pigs kept in the smaller groups (Nielsen et al., 1995), clearly showing that their feeding strategy was also changed with group size. Results from free-range conditions suggests that group size also may depend on the activity of the pigs: pigs tend to split into smaller subgroups during the day when they are foraging, while they are united as larger groups during rest at night (Rodrìguez- Estèvez et al., 2010). This may be a good anti-predator strategy for the pigs under free-range conditions but could also very well reflect similar behavioural needs under indoor conditions. However, there is little evidence for more subgrouping in larger groups of pigs kept indoors but rather that they tend to be more dispersed than in smaller groups (Turner et al., 2003). It is thus likely that sub-grouping relates to resource distribution rather than the group size per se. One possible, negative side effect of larger groups with more total space for instance on deep straw bedding could be that weight gain might be reduced compared to smaller group pens or that it demands a higher feed consumption to keep a similar weight gain (Turner et al., 1999). A reason for higher feed consumtion can be an increased activity in larger groups with more total space, as it has been documented that pigs from large groups (80 pigs) in pens with deep straw bedding are standing more, show more locomotion and interact more with their environment (i.e. social interactions and exploration) than pigs in conventional, smaller pens with a group size of 15 (Morrison et al., 2006). However, nor the latter study or older studies (Nielsen et al., 1995), find differences in growth performance between group sizes, suggesting that other factors in the environment may interact with group size and pen system. Meat quality data shows that loins from the deep-litter pens with large groups also had a lower pH, more purge loss, more glucose in purge and were lighter in subjective color. On the other hand, there was no difference in tenderness, juiciness or overall desirability detected by a trained sensory panel. We do not yet know what the optimal group size for pigs is regarding optimizing social behavior, welfare, growth performance and meat quality, but there are many welfare benefits of larger group pens as it is more stimulating for the pigs. Effects of group size needs to be studied systematically in combination with for instance pen design, animal density, feeding system, environmental enrichment, and resource distribution (i.e. number and location of feeders and drinkers, rooting material, as well as attractive resting places) as these factors most likely will interact. Our current field survey and project on Norwegian fattening pig welfare and productivity will cover all these factors.

Regarding animal density, finishing pigs (around 75 kg) housed at 0.8 m² show more negative social behavior, has more lesions on all parts of the body including ears, are less clean and has a higher body surface temperature than those housed at space allowances of 1.2 or 1.6 m² per pig, respectively (Fu et al., 2015). Interestingly, the number of positive social behaviours in this study was greatest in the intermediate density. Comparatively, Norwegian legislations state a minimum space requirement of 0.65 m² for pigs in this weight category (www.Mattilsynet.no) and could be considered too small to ensure good welfare conditions. Which is the most important factor for pig welfare, space or enrichment? Beattie et al. (1996) attempted to answer this question many years ago by comparing the following space allowances for newly weaned groups of 6-week-old pigs: 0.5, 1.1, 1.7 or 2.3 m²per pig, all being enriched with free access to the substrates peat and straw, and added a fifth group with the largest space allowance but with no enrichment. The results showed that there was less exploration of substrates and more inactivity in in the enriched group with the smallest space allowance and the non-enriched group with the largest space allowance. Pigs moved more when given greater floor space in enriched pens, but duration of harmful, social behaviour was greater in the barren treatment with the largest space allowance than in the four enriched treatments with different space allowances, suggesting that enrichment is more important than space allowance per se but that a minimum amount of space is needed for the pigs to be able to explore and interact in a positive way. From this study it was concluded that enrichment played a greater role in determining pig behaviour than floor space allowance, which makes sense since the pigs motivation to move around and engaging in positive activities are goal directed. For instance, pigs explore to find food and chewable objects, and space to move around in itself is not enough to satisfy that need. However, to be able to offer more optimal environments for the entire growing and finishing period, we need to combine both physical and social enrichment (i.e. larger groups, group composition, stable littermate groups) with a wider range of space allowances and pen designs as resource distribution, location of rest areas etc. are also likely to affect this picture to a large extent. If pigs could work to get access to attractive resources such as a self-administered rooting material dispenser, pigs would be even more stimulated than when giving them free access. Overall, there is not enough scientific documentation of the effects of space allowance in combination with factors such as source and frequency of enrichment and pen design.

Digital technology and deep learning methods for automatic recognition of pig behaviours – a powerful tool for animal welfare assessment

New digital technology to recognize positive and negative behavioural indicators will become an important welfare assessment tool in the future. In our recent project “DigiPig,” our goal is to produce a digital surveillance system for behavioural recognition of pigs based on video sequences and machine learning, combined with an “app” where the farmer can keep track of a welfare protocol as well as management routines to assess welfare status at the farm over time, and to have immediate feedback on his or hers everyday routines. The concept will include both welfare and productivity measures, and the goal is to increase and simplify the everyday awareness of the condition of animals in the herd, and to give the farmer an efficient tool for this purpose. Making online animal welfare courses is a good thing, but in the end, it is the farmer's everyday practical routines and handling of the animals that are going to make a difference. Therefore, we must develop a system that is motivating and user friendly for the farmer, at a low cost, work efficient, and leading to a welfare friendly, sustainable animal production. Based on annotating 600 images from 2D video recordings of groups of weaned pigs provided with rooting substrate, we have developed a program that recognizes each individual pig with a precision of 96%, the tail with 77% and heads with 66%, respectively (Figure 2). Surprisingly, the tail was easier to identify than the head. This is also a positive result for the future development of the system, as we predict that a straight tail is associated with a negative affective state, a curled tail with a neutral to positive state and a wagging tail to an excited (aroused), positive state. This was also demonstrated in our recent study on weaned pigs (Ocepek et al., under review). The tail is thus considered one of the major indicators of the welfare conditions of young pigs. However, as we would like to have a screening of the situation both on individual and group level, we consider the following behaviours to be important:

Pig tail: straight hanging down, curled, wagging
Individual deviation from behavioural synchrony (as pigs in a group usually have a synchronous activity pattern (i.e. eating, resting etc.)
Solitary and social play
Exploration
Fighting
Ear and tail biting
Pig vocalisations: Low-frequent grunts vs. pig screaming

Most of these behaviours are associated with a certain body posture that can be recognized in the similar way as the sow behaviours in the recent studies mentioned below. Deviation from synchrony and vocalisations can be done on group/pen level whereas the others are on individual level.

Several sensor modalities are now available for automatic monitoring of behaviour, for instance deviations in drinking and feeding, frequency of coughs and vocalisations have been registered by using such systems combined with automated alerts sent to mobile phones (Matthews et al., 2016). Deviations from behavioural synchrony in groups of pigs could also potentially be interesting as they tend to show very synchronous activity patterns and the individuals deviating from this pattern could potentially have some problems. Similarly, sound recording on group level could reveal if pigs are screaming to a large extent or if the group sound is dominated by low-frequent grunts associated with a positive mental state. Programs using facial recognition of individual pigs show an accuracy of 96.7% from 1553 images (Hansen et al., 2018). However, if one wants to monitor several pens with many pigs in each and get an insight into their welfare status, we need cameras from above with a slight side angle, and thus it would be easier to recognize pigs based on shapes rather that faces as their face are usually oriented towards the ground. The face will thus be less visible than their body posture or shape, tail or ears.

Figure 2. The annotations were created with Labelbox which is a customizable collaborative online tool. Labelbox is free for academic research. The image material was extracted from 400 hours of video.

The annotations were created with Labelbox which is a customizable collaborative online tool. Labelbox is free for academic research. The image material was extracted from 400 hours of video.

Deep learning and machine vision approach for posture detection of individual pigs has great potential as a welfare assessment tool. Two-dimensional imaging system along with deep learning approaches can successfully be utilized to detect the standing and lying (belly and side) postures of pigs under commercial farm conditions (Nashiramadi et al., 2019). Data from different commercial farms were used for training and validation of the proposed models. Experimental results show that for instance the R-FCN ResNet101 method was able to detect lying and standing postures with a mean precision of more than 93%. This is extremely interesting as both positive behaviours, such as play and exploration, and negative behaviours such as in aggressive conflicts are associated with certain postures that most likely can be recognized from images. Similarly, Yang et al. (2018 a), used deep learning for automatic recognition of sows’ nursing behaviours in 2D images, with an accuracy of 97.6%, sensitivity of 90.9% and specificity of 99.2%. Faster R-CNN and ZFnet has been applied to recognize individual feeding behaviours of pigs (Yang et al., 2018 b), where each pig in the barn was labelled with a letter. Their proposed method was able to recognise pigs’ feeding behaviours with an accuracy of 99.6%.

Image analysis techniques using fully convolutional networks (FCNs), appears to be one of the most promising methods of automatic recognition of sow behaviours from video sequences. In a study of lactating sows (Yang et al., 2020), temporal features that evaluated the temporal motions of the animals were extracted, and these spatial and temporal features were then put into a hierarchical classifier for behavioural recognition. Based on 468,000 frames of 3 sows, accuracies of behavioural classification compared to manual scoring was: 97.49% for drinking, 95.36% for feeding, and 88.09% for nursing, respectively. We will look into the details of this method for the further development of DigiPig.

Eliminative pig habits and pen design – how can we achieve a dry and a clean pen?

Understanding eliminative (urination and defecation) pig habits is important to achieve a dry and clean pen. Inappropriate eliminative behaviour causes fouling of pen resting (lying) area, and this has a negative effect on the environment, for instance regarding ammonia emission (Ocepek and Škorjanc, 2016), the cleanliness of pigs and pens (Andersen and Pedersen, 2011; Bøe et al. 2019), pig and human health (Urbain et al., 1994).

Pigs are some of the cleanest animals as they distinguish between areas for resting and eliminating (Whatson, 1978; Stolba and Wood-Gush, 1989). For resting areas, pigs prefer to choose an area with a solid floor (Aarnink et al., 1997), in close proximity to feeders (Baxter, 1982), without disturbance from other pigs (e.g. the neighbouring pens; Hacker et al., 1994), and with a warm surface (Marx and Buchholz, 1989). To keep the resting area clean, pigs choose to eliminate as far away as possible from it (Stolba and Wood-Gush, 1989; Ocepek and Škorjanc, 2016), and in a separate area when this is available (inner vs. outer; Guo et al., 2015; Ocepek and Škorjanc, 2016; Ocepek et al., 2018). They prefer to eliminate on slatted floor (Aarnink et al., 1997), in cold (Banhazi, 2013), bright (Taylor et al., 2006), wet (Baxter, 1982) areas, in close proximity to walls or in the corner of their pens (Baxter, 1982; Bate et al., 1988). Pigs have also been observed to eliminate close to drinkers (Ocepek et al., 2018). The placement of drinkers in the outside area compared to the inside area resulted in more than a 30% decrease in the eliminating on resting area and a 20% increase in eliminating on slatted floor, thereby reducing fouling, less time needed for manual cleaning, and improved pig welfare. This information is of great interest if we want to plan future housing systems with an adjacent outdoor area. If the outdoor area makes it easier to clean because of increased urination and defecation on the slats inside, this could also be an important economic argument for motivating farmers to use outdoor areas as it decreases the heavy workload of cleaning pens.

Pigs’ eliminative behaviour can be altered by a few environmental factors. At higher ambient temperatures (≥19 °C), pigs begin seeking cooler areas for resting (Huynh et al., 2005). Over solid floor, pigs start to prefer slatted area for resting as it is cooler (Huynh et al., 2004). Thus, we should aim to keep room temperature below this limit, and rather make warmer microclimates for young pigs that have lower critical temperature. At a higher density (less lying space per pig), pigs eliminate more frequently on solid floor, while they begin to rest more frequently on slatted floor areas (Ocepek and Škorjanc, 2013; Ocepek and Škorjanc, 2016). Reducing animal density in the pens is therefore crucial. We can achieve a dry and a clean pig pen by providing an optimal resting area and area for eliminating. Resting area should consist of solid, insulated floors, closed pen partitions, and feeders located in the corners. Lying areas must be large enough for pigs to rest comfortably during the whole growing-fatting period. By contrast, eliminating area should be a separate area, preferably outdoor, on slatted and wetted floor, bright enough and cooler, with access to drinkers.

Conclusion - important criteria for future housing design

It is beneficial to keep as stable social groups as possible and to design an environment where attractive resources, such as drinkers, feeders, areas for rooting or resting are distributed in a way that makes it difficult to monopolize for a few individuals.

Pigs should be offered more space and substantially more enriched environments including different sources of rooting material at least twice a day and positive challenges (i.e. self-administered rooting material dispenser or other simple problem-solving tasks) in their environment in all stages of production from birth until slaughter. This would lead to less frustration and therefore fewer incidents of redirected, harmful behaviours and rather an increase in positive, social behaviours.

Planning of future housing systems should include greater use of controlled, outdoor arenas. This is a cheaper way of making more stimulating environments, and results in more optimal eliminative habits such as using the inside, slatted area more. This will also reduce the workload regarding cleaning of pens.

Larger groups with a larger total space are socially more stimulating, leads to more locomotion and a higher activity level, and demands less pen equipment and pen dividers. Deep bedding can be removed once or twice a year with a tractor, whereas most small pen systems require substantial effort every day to clean each pen.

To achieve a sustainable pig production in the future, there is a need for substantially more complex pig environments than what we see in most pig production units today. This requires a change is how we value meat, and that the consumers are willing to pay more for good quality meat, where quality also means “quality of life” for these animals. This is also needed to ensure good human health in the future.

Published in the proceedings of the International Pig Veterinary Society Congress – IPVS2020. For information on the event, past and future editions, check out https://ipvs2022.com/en.

Understanding the behavioural needs of growing and finishing pigs and how we can meet those needs when designing future housing systems

Antibiotics reduction in pigs: challenges and opportunities

Potassium diformate as acidifier