Applied Animal Behaviour Science 241 (2021) 105383 Contents lists available at ScienceDirect Applied Animal Behaviour Science journal homepage: www.elsevier.com/locate/applanim The potential of feeding patterns to assess generic welfare in growing-finishing pigs Jacinta D. Bus *, Iris J.M.M. Boumans, Laura E. Webb, Eddie A.M. Bokkers Animal Production Systems group, Wageningen University, P.O. Box 338, AH 6700 Wageningen, the Netherlands A R T I C L E I N F O A B S T R A C T Keywords: The welfare of growing-finishing pigs is important inherently to the pigs, but also for the societal acceptance and Feed intake environmental impact of the husbandry system. Nevertheless, methods to monitor pig welfare throughout the Feeding frequency whole growing-finishing phase have not yet been successfully developed. One possibility is to use electronic Feeding duration feeding stations to identify the feeding pattern of individual pigs, which consists of feed intake and the behav­ Feeding rate Animal welfare iours underlying intake (i.e. feeding frequency, duration and rate), and to process this data using an algorithm Sensor technology that links feeding patterns to pig welfare. Before such a monitoring system can be developed, a thorough un­ derstanding of both pig feeding patterns and their relationships with pig welfare is required. The aim of this review was to assess the current state of this understanding. We begin with a narrative review that describes the feeding patterns of growing-finishing pigs, and subsequently provide a systematic review of the relationships between pig feeding patterns and welfare. We focused on animal-based parameters of pig welfare, but also included resource-based parameters known to influence welfare (e.g. space allowance, environmental enrich­ ment). We found that so far, studies have focused on physiological and behavioural welfare problems, while the affective part of welfare, both positive and negative, has been largely overlooked. Deviations from basal feeding patterns may occur during reduced welfare states, sometimes even preceding other clinical or behavioural manifestations of the problem. Particularly clear are the links between feed intake and physiological causes of reduced welfare, such as clinical health, thermal stress and tail biting wounds. The behaviours underlying intake provide further information, as they show distinct deviations in response to different physiological welfare problems and as their rapid responses may enable detection of disease at a subclinical stage. However, a wider range of clinical diseases should be studied before this knowledge can be applied. Behavioural welfare problems, such as abnormal behaviours and feed competition, mostly induce deviations in the feeding behaviours under­ lying intake but not intake itself, though more knowledge is required to confirm this finding. We conclude that feeding patterns are a promising tool to monitor generic pig welfare. Feed intake and the behaviours that un­ derlie it should be used simultaneously, on a short time scale (i.e. within the day). It should be considered that the variation in feeding patterns between and potentially within pigs is large, and that this variation should be well-understood before welfare-relevant variation can be interpreted. 1. Introduction types of pigs is additionally specified), welfare is mostly monitored by the farmer, who identifies clinical disease or behavioural problems and 1.1. Background intervenes. With the long-standing increase in farm sizes (CBS, 2015), it becomes increasingly difficult for farmers to obtain a good picture of the The welfare of farm animals (hereafter referred to as animals) is welfare of each pig under their care. In addition, farmers are likely to important not only inherently for the animals, but also for the societal only identify clinical disease and obvious behavioural deviations that acceptance (Broom, 2010) and environmental impact (Chemineau, require intervention, while other causes of reduced welfare, such as mild 2016) of husbandry systems. In growing-finishing pigs (for simplicity, social stress or subclinical disease, may go unnoticed. To obtain a more ‘pig’ is used to refer to growing-finishing pigs, while reference to other detailed and complete picture of pig welfare, an external assessment, * Corresponding author. E-mail address: jacinta.bus@wur.nl (J.D. Bus). https://doi.org/10.1016/j.applanim.2021.105383 Received 6 March 2021; Received in revised form 16 June 2021; Accepted 18 June 2021 Available online 25 June 2021 0168-1591/© 2021 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). J.D. Bus et al. Applied Animal Behaviour Science 241 (2021) 105383 such as the Welfare Quality® protocol (Welfare Quality®, 2009), can be “cited by”-function of a search engine, here Google Scholar. used. Although external assessments give more detail and provide in­ Search 1) used a search string consisting of three parts: “feeding formation that could be communicated to internal and external parties, pattern” AND “growing-finishing pigs” NOT “non-target animals”. The such as farmers, farm advisors and consumers, they only provide a exclusion was solely used to reduce the number of irrelevant results. The snapshot in time and are very time consuming to perform. In addition, exact search string was: ("feed intake" OR "feeding duration" OR "feeding they often suffer from complex, unbalanced and subjective aggregations frequency" OR "feeding time" OR "feeding visit*" OR “feeder visit*” OR of different aspects of welfare to come to an overall welfare score "feeding pattern" OR "feeding behavio*r") AND ("growing-finishing pig*" (Czycholl et al., 2017; de Vries et al., 2013). OR "growing pig*" OR "fattening pig*" OR gilt* OR barrow* OR boar* An alternative method comes from precision livestock farming (PLF), OR "meat pig*" OR grower* OR finisher*) NOT (broiler* OR cattle OR which could potentially enable continuous monitoring of (individual) sheep OR goat* OR fish). From the resulting articles, only peer-reviewed pigs to rapidly identify reduced welfare states and gain knowledge on articles in English, Dutch and German were selected, and reviews, meta- the welfare of each pig across its entire growing-finishing phase analyses, reprints, MSc theses, articles without full-text access and ar­ (Berckmans, 2014). For example, microphones can be used to detect ticles concerning other types of pigs or animals than growing-finishing respiratory disorders at group level (Guarino et al., 2008), and 3D pigs were excluded. For the remaining articles, relevance for inclusion cameras have been used to predict tail biting outbreaks by automatically in the systematic review was judged based on title, abstract and full-text identifying lowered tail postures (D’Eath et al., 2018). Another prom­ relevance, in that order. Relevant articles included at least one param­ ising type of PLF technology monitors the feeding patterns of individual eter of feeding behaviour and compared this to either 1) animal-based pigs, using electronic feeding stations, radio-frequency identification parameters of pig welfare or 2) resource-based parameters that are (RFID) antennas at the feeder or computer vision, although the latter known to influence pig welfare, such as space allowance or environ­ two provide less detail than electronic feeding stations. As numerous mental enrichment. Backward and forward snowballing were performed relationships between feeding patterns and welfare aspects have been on the resulting articles to identify additional articles, based on title reported (e.g. with disease (Brown-Brandl et al., 2013; Munsterhjelm relevance. Subsequently, these articles were subjected to the same in­ et al., 2015) and thermal stress (Brown-Brandl et al., 2000; Nienaber clusion criteria as during the original search. et al., 1996)), these feeding patterns could potentially be used to To ensure that the first wide search had identified all papers relating monitor pig welfare. Nevertheless, a thorough understanding of both pig (aspects of) feeding behaviour to pig welfare, and collect any articles not feeding patterns themselves and their relationships with pig welfare yet obtained, two additional, narrower searches were performed (2a & should be available before that potential can be utilised. 2b). The search strings included the same three parts as during the first search, but added a fourth part that included 2a) a range of clinical 1.2. Aim of the review diseases or 2b) a range of keywords for negative and positive welfare not related to clinical disease. The strings used were: 2a) AND (health OR The aim of this review was to assess the current state of our under­ disease OR injury OR pneumonia OR lameness OR lame OR bursitis OR standing of pig feeding patterns and their relationships with pig welfare. diarrhoea OR diarrhea OR clinical OR subclinical OR pathogen OR This review consists of two parts: 1) a narrative review describing lesion OR lesions OR hygiene OR cleanliness OR sanitary OR coughing feeding patterns in pigs and their variation, and 2) a systematic review of OR sneezing OR pumping OR scouring OR wound OR wounds); and 2b) the relationship between feeding patterns and several aspects of pig AND (“thermal stress” OR “heat stress” OR “cold stress” OR “high tem­ welfare. In the next section, we first describe the steps made to obtain perature” OR “low temperature” OR “social stress” OR dominance OR and select the relevant literature. This is followed by the narrative and “rank order” OR competition OR “social facilitation” OR “tail biting” OR systematic reviews and, finally, a discussion integrates the acquired “abnormal behavio*r” OR “belly nosing” OR “positive welfare” OR information and assesses whether a good coverage of the different as­ “positive emotion” OR “positive mood” OR enrichment OR enriched OR pects of pig welfare has been obtained. straw). The resulting articles were processed using the same criteria as those of the first search, and forward and backward snowballing were 2. Methods performed in identical fashion. Only articles that had not yet been identified in the first search were included, otherwise these are marked This review consists of a narrative and a systematic literature search as duplicates. that were both completed by the first author on June 17th, 2020, and In total, 116 papers were included in the systematic review. An includes relevant articles published before that date. For both the overview of the number of papers considered at each step of the selec­ narrative and the systematic review, the same searches were performed. tion process can be found in Fig. 1. Based on the contents of the included However, for the systematic review the papers were processed and papers, six categories of welfare parameters were identified: clinical included systematically (n = 116), while for the narrative review a se­ disease, subclinical disease, thermal stress, social behaviour, tail biting lection of the most relevant papers (n = 100) was made due to the high and environmental enrichment. In the systematic review, we discuss the number of papers reporting similar results. The systematic part, which literature for each welfare category separately. Each category also in­ focused on the relationship between feeding patterns and pig welfare, cludes a table that summarises the results for the main feeding param­ used primarily animal-based indicators of welfare, and included eters: feed intake, and feeding duration, frequency and rate. Note that resource-based indicators if these were known to be relevant for pig not all studies included in the systematic review may be present in these welfare, such as space allowance or environmental enrichment. Welfare summary tables, as some were of a design that was not relevant for in­ was defined here as the balance between positive and negative experi­ clusion, such as modelling studies or results of behavioural tests. Most of ences (e.g. Green and Mellor, 2011; Webb et al., 2019). Three searches the included studies observed feeding patterns in the home pen under ad were performed: 1) a wide, general search to identify papers on pig libitum feeding, but some used restricted feeding or a behavioural test. A feeding patterns, and two narrower searches focused on either 2a) the full overview of all articles included in the systematic review and their relation between clinical health and pig feeding patterns or 2b) the findings can be found in the Supplementary materials. relation between other aspects of negative or positive welfare and pig feeding patterns. All searches were performed using Topic search terms 3. Feeding patterns in pigs in Web of Science, and backward and forward snowballing were per­ formed on the selected articles. Backward snowballing uses the refer­ 3.1. A description of pig feeding patterns ence list of a source article to identify additional articles, while forward snowballing identifies articles that cited the source article using the Pig feeding patterns consist primarily of feed intake and the 2 J.D. Bus et al. Applied Animal Behaviour Science 241 (2021) 105383 Fig. 1. A flow chart of the procedure of article selection for the systematic review, including the number of articles considered at every step. behaviours that underlie intake: feeding frequency, duration and rate. meal. However, statistical methods to calculate meal criteria vary Behaviourally, feeding rate is a driver of feed intake, but in practice it widely between authors and their biological relevance (i.e. what is the can only be estimated by dividing intake by duration. These parameters pig doing between visits?) is not always clear (reviewed by Maselyne can be expressed on multiple time scales, such as per 24 h, per feeding et al., 2015). In addition to the basic parameters, the frequency of visit or per meal. A visit is generally defined as every time a pig enters non-nutritive visits (i.e. when the pig visits the feeder but does not the feeder, while a meal is defined as an accumulation of sequential consume any feed) (Garrido-Izard et al., 2020; Young and Lawrence, visits that were separated by an interval below a calculated meal crite­ 1994) and the interval between visits or between meals (Herrer­ rion (Howie et al., 2009; Maselyne et al., 2015). The underlying a-Cáceres et al., 2019; Kapun et al., 2017, 2016; Nienaber et al., 1991) assumption is that between rapidly successive visits the pig remains are parameters contributing to feeding patterns. Some authors have also mentally occupied with feeding, as satiety has not yet been reached, and included additional parameters, such as time spent eating simulta­ that, therefore, these visits should be considered as one feeding event: a neously with other pigs (Kapun et al., 2017) or time spent queuing for 3 J.D. Bus et al. Applied Animal Behaviour Science 241 (2021) 105383 access to the feeder (Morrow and Walker, 1994a; Rasmussen et al., 2006a, 2006b; Walker, 1991). The different feeding parameters are, as could logically be expected (Nielsen, 1999), strongly related to each other, where correlation co­ efficients of above 0.8 are repeatedly reported (Carcò et al., 2018; Von Felde et al., 1996). For example, at pig level it has been reported that a higher feed intake relates to a longer eating time (Guo et al., 2015b; Hyun and Ellis, 2002, 2000; Rauw et al., 2006; Schulze et al., 2001), that a lower feeding frequency relates to a higher intake and duration per visit (Carcò et al., 2018; Fernández et al., 2011; Garrido-Izard et al., 2020; Hall et al., 1999a; Hyun et al., 1997; Hyun and Ellis, 2002, 2000; Nielsen et al., 1995; Schamun and Hoy, 2011; Schulze et al., 2001; Young and Lawrence, 1994); and that a higher feeding rate relates to a higher feed intake per day and per visit and a lower daily feeding Fig. 2. The alternans diurnal pattern in feed intake shown by group-housed duration (Carcò et al., 2018; Chen et al., 2010; Garrido-Izard et al., growing-finishing pigs, based on de Haer and Merks (1992), Schouten (1986) 2020; Guo et al., 2015b; Herrera-Cáceres et al., 2019; Hyun and Ellis, and Montgomery et al. (1978). Note that the exact timing of the peaks depends on the lighting regime applied, and that the results are shown at group level, 2002, 2000; Ragab et al., 2019; Schulze et al., 2001). These relationships averaged over time. between feeding parameters appear to have a genetic basis (Ding et al., 2018; Herrera-Cáceres et al., 2019; Labroue et al., 1997; Rohrer et al., 2013; Von Felde et al., 1996), and indicate that individual pigs have pattern is seen for feed intake and feeding duration per visit (Bigelow distinct behavioural strategies to achieve their desired nutritional intake and Houpt, 1988; Chen et al., 2010; Hyun et al., 2001, 1997; Hyun and (Fernández et al., 2011; Garrido-Izard et al., 2020). In support of the Ellis, 2001). This can be explained by a high level of competition at the latter, it was reported that feeding patterns recorded at the beginning of feeder during peak hours, which leads to frequent displacements and the growing period correlated with those recorded at later stages on the thus many short feeding visits. Indeed, when the number of pigs per same animals (Ragab et al., 2019; Rauw et al., 2006). feeder and hence competition is low, no lower levels in intake and Nevertheless, feeding patterns are known to change with age. Many duration per visit during peak hours were seen (Hyun and Ellis, 2002). studies have reported that daily feed intake increases with body weight Feeding rate is reported to be quite constant throughout the day (Bau­ in a linear fashion, until a plateau in intake is reached (Casey, 2003; mung et al., 2006; Feddes et al., 1989; Hyun et al., 2001), although it Fàbrega et al., 2003; Hauschild et al., 2020; Kavlak and Uimari, 2019; might increase during the afternoon peak (Andretta et al., 2016; Bau­ Labroue et al., 1999; Lorenzo Bermejo et al., 2003; Mensching et al., mung et al., 2006; Feddes et al., 1989; Fernández et al., 2011; Garri­ 2018; Morgan et al., 2000; Nielsen et al., 1995; Putz et al., 2019; Rauw do-Izard et al., 2020; Young and Lawrence, 1994). This implies that et al., 2006). When exactly this plateau is reached can vary largely, with feeding rate is less easily adapted during competition than feeding fre­ reported ages differing from 99 to 180 days (Fàbrega et al., 2003; quency and duration. It should be noted that all patterns in feeding Godcharles et al., 1993; Lorenzo Bermejo et al., 2003). Although behaviours strongly depend on Zeitgebers, which are environmental increasing feed intake is initially mediated by increasing daily feeding factors that synchronise the internal circadian clock to real day length duration (Hyun and Ellis, 2000; Rauw et al., 2006), feeding duration (Aschoff, 1966), such as the light-dark cycle, feeding times and caretaker seems to later decrease with age (Hall et al., 1999b; Hyun and Ellis, activity. Of all possible Zeitgebers, the lighting regime appears partic­ 2000; Kavlak and Uimari, 2019; Labroue et al., 1999; Mensching et al., ularly important, as application of continuous lighting causes the 2018; Rauw et al., 2006). In these later stages, the increase in feed intake diurnal feeding pattern to collapse (Hsia and Wood-Gush, 1984a; is achieved through an increase in feeding rate (Hall et al., 1999b; Ingram et al., 1980), especially in individually-housed pigs (Ingram Kavlak and Uimari, 2019; Labroue et al., 1999; Mensching et al., 2018; et al., 1980). Nevertheless, a weakened form of the diurnal pattern may Quiniou et al., 2000a). Feeding rate increases because body size and oral still be maintained during continuous lighting if other Zeitgebers are capacity become larger over time, which allows pigs to take larger bites present (Hyun et al., 1997). and eat more food in less time. Indeed, if feeding rate is expressed per kilogram of metabolic weight, it is seen to decrease rather than increase 3.2. Variation in feeding patterns over time (Nienaber et al., 1991, 1990). Feeding frequency is generally reported to reduce over time (Hoy et al., 2012; Labroue et al., 1999, It has repeatedly been acknowledged that the variation in feeding 1994; Quiniou et al., 2000a; Walker, 1991), while intake (Hall et al., patterns, both within and between studies, is large. This is mainly the 1999b; Hoy et al., 2012; Kavlak and Uimari, 2019; Mensching et al., case for the underlying feeding behaviours, while variations reported for 2018; Nienaber et al., 1990; Quiniou et al., 2000a) and duration (Nie­ daily feed intake are much smaller (Ding et al., 2017). For example, the naber et al., 1990; Walker, 1991) per visit or meal are reported to in­ standard deviations of feeding frequency, duration and rate commonly crease over time. This change to larger meals is likely a consequence of reach values as high as a third of the group mean (Ding et al., 2017; Do enlarging gut capacity (Boumans et al., 2015; Quiniou and Noblet, et al., 2013b; Jiao, 2015; Maselyne et al., 2014). Moreover, only 16 % of 2012). animals were reported to show a daily feeding duration within one If we look at the pattern of feeding behaviour within a day, under ad standard deviation of the mean (Andretta et al., 2016), and for all pa­ libitum feeding, an alternans pattern in feed intake can be identified (e.g. rameters pigs regularly showed values more than three standard de­ De Haer and Merks, 1992; Montgomery et al., 1978; Schouten, 1986). viations away from the mean (Reyer et al., 2017; Young and Lawrence, An alternans feeding pattern is characterised by a majority of feed intake 1994). Similarly, the alternans pattern may vary strongly, with some occurring during the day, with a small peak in the morning and a larger studies reporting that the morning peak was larger than the afternoon peak in the afternoon (Fig. 2). Although this pattern is still weakly peak (Hoy et al., 2012; Hyun and Ellis, 2002, 2000), while others report developed in young pigs, it becomes more distinct with age (Morrison that pigs had only one peak in feeding activity, either in the morning et al., 2003; Toft and Madsen, 1998; Weiler et al., 2013). The peaks in (Augspurger et al., 2002; Hyun and Ellis, 2001) or in the afternoon feed intake coincide with peaks in feeding duration (e.g. Laitat et al., (Baumung et al., 2006; Fraser et al., 1991), or that there was a high 2015; Renaudeau, 2009; Villagrá et al., 2007; Weiler et al., 2013) and feeding activity throughout the day (Weiler et al., 2013). Studies that frequency (e.g. Fernández et al., 2011; Garrido-Izard et al., 2020; Hoy looked at the feeding patterns of individual pigs noted that although et al., 2012; Xin et al., 2016; Young et al., 2011), while the opposite many pigs show an alternans feeding pattern, others may eat throughout 4 J.D. Bus et al. Applied Animal Behaviour Science 241 (2021) 105383 the day (Gertheiss et al., 2015), may show no consistent pattern across intake, called anorexia, as part of the sickness response (reviewed by days (Carcò et al., 2018; Hyun et al., 1997; Maselyne et al., 2018; Toft Johnson, 2002). In pigs, several studies have reported such an intake and Madsen, 1998), or may even mainly feed at night (Ingram et al., drop, for example during pneumonia (Brown-Brandl et al., 2013), 1980). lameness (Munsterhjelm et al., 2015), bacterial infection (Helm et al., It is likely that part of this variation can be explained by differences 2018a, 2018b), and viral infection (Schweer et al., 2016), although one in study methodologies, such as the way in which feeding behaviour is study reported no reduction in feed intake during mild viral infection (Er measured (e.g. using electronic feeding stations, RFID antennas at et al., 2014). Deviations in daily feed intake have been used to effec­ feeding troughs, or scan sampling from video footage) and processed (e. tively identify more and less disease resilient pigs, where the authors g. including or excluding non-nutritive visits), or in which time frame it defined disease resilience as the ability to maintain relatively undi­ is expressed (e.g. hourly, daily, or averaged across the growing period or minished performance during infection (i.e. pigs more affected by dis­ in smaller periods). Beyond that it is, however, known that character­ ease would show larger reductions in feed intake) (Putz et al., 2019). istics of individual pigs and environmental circumstances explain part of Nevertheless, as clinical disease was not monitored, in the study of Putz the variation, such as sex and castration (e.g. Fàbrega et al., 2010; Guo et al. (2019) it cannot be reliably concluded that pigs with deviating et al., 2015a; Schmidt et al., 2011), breed (e.g. Augspurger et al., 2002; feeding patterns were indeed less resilient to disease, although at group Fernández et al., 2011; Renaudeau et al., 2006, 2005), genetics (e.g. level resilience was seen to correlate with veterinary treatments and Ding et al., 2017; Do et al., 2013b; Guo et al., 2015b), body weight and mortality. Feed intake reduction may occur long before clinical signs are growth rate (e.g. Brown-Brandl et al., 2013; Gonyou and Lou, 2000; evident, such as up to twenty days before diagnosis of lameness (Mun­ Turner et al., 2002), personality (e.g. Boumans et al., 2018a; Rohrer sterhjelm et al., 2015), but in the case of infection it usually occurs after et al., 2013), feed content (e.g. Castilha et al., 2016; Quemeneur et al., initial infection (Sandberg et al., 2006). The severity of the intake 2019; Whittemore et al., 2001a; Zeng et al., 2019), the use of dry or wet reduction can be highly variable. Munsterhjelm et al. (2015) reported feed (e.g. Bergstrom et al., 2012; Botermans and Svendsen, 2000; that the extent of the daily feed intake reduction in response to lameness Gonyou and Lou, 2000), the type and location of the feeder (e.g. depended on pig age, with younger pigs being more affected, and Botermans et al., 2000; Botermans and Svendsen, 2000; Nielsen et al., differed between pigs that would or would not recover (25.5–45.6 % vs. 1996a), lighting regime (e.g. Feddes et al., 1989; Ingram et al., 1980), 55.3–99.2 % reduction in intake). There are also some indications that season (e.g. Mensching et al., 2018; Schulze et al., 2001; Xin et al., intake reduction depends on sex and genetic line (Frank et al., 2005), 2016), management procedures such as weighing (e.g. Augspurger and and it has been theorised that onset, severity, and recovery differ be­ Ellis, 2002), and the lasting effects of the environment and condition of tween diseases (Munsterhjelm et al., 2015), pathogen types (Ahmed the mother during gestation (e.g. Sell-Kubiak et al., 2013). In addition, et al., 2014) and pathogen load during initial infection (Frank et al., some of the variation may be due to the differences in pig welfare states. 2005). It is generally assumed that pigs do not compensate for reduced For example, clinically diseased pigs will show a lower feed intake than intake by increasing intake post-recovery (Pastorelli et al., 2012), healthy pigs (e.g. Brown-Brandl et al., 2013; Helm et al., 2018a, 2018b; however, there is no empirical evidence supporting or refuting this Munsterhjelm et al., 2015; Schweer et al., 2016), and pigs subjected to assumption. One study has reported that feed intake increased after severe feed competition will show lower feeding frequencies and higher antibiotic administration (Hauschild et al., 2020), but as the reason for intakes per visit than pigs housed in non-competitive situations (e.g. administration was not reported, no conclusion regarding compensatory Georgsson and Svendsen, 2002; Labroue et al., 1999; Morrow and feeding can be drawn from this finding. Walker, 1994a; Nielsen et al., 1995; Wallenbeck and Keeling, 2013). In Although the impact of clinical disease on feed intake is quite clear, the next section, we describe these relationships between feeding pat­ there is less knowledge on its impact on the behaviours that underlie terns and pig welfare in detail, using a systematic literature review. intake. Several studies have reported that daily feeding duration severely reduces during clinical disease (Adrion et al., 2018; Ahmed 4. Feeding patterns and pig welfare et al., 2014; Kapun et al., 2016; Putz et al., 2019), leading Brown-Brandl et al. (2016) to use modelling techniques to detect reductions in feeding 4.1. Clinical disease duration to identify pigs suffering from pneumonia. With their model, they were indeed able to detect all 5 sick pigs and 17 out of their 21 sick On the relation between feeding patterns and clinical disease, 18 days, but as these pigs were initially identified as suffering from pneu­ papers were identified. A summary of the main findings is provided in monia by their large drop in feed intake, not from clinical signs, they Table 1. represent only unconfirmed cases. In addition, it is likely that only se­ Across animal species, clinical disease induces a drop in daily feed vere cases were detected here, as less severe cases could be theorised to Table 1 The effect of clinical disease on feeding patterns (intake (inta), duration (dur), frequency (freq) & rate) compared to feeding patterns of healthy pigs (increase ↑, decrease ↓, or no difference ≈). Empty cells show that no papers were found on this topic. The final column indicates the number of papers that support or contradict this finding. An overview of the papers this table is based on can be found in Supplementary Table 1. Daily level Meal level Visit level Reference count Clinical disease Inta Dur Rate Inta Dur Freq Inta Dur Freq Support Contrast Coughing ↓ 1a 0 Diarrhoea ≈ 1a 0 Lameness ↓ ↓ 2a 0 Osteochondrosis ≈ ↑ ↑ ↓ 1 0 Pneumonia ↓ 4a 0 Skin lesions ↓ ≈ 2a 0 Bacterial or viral infection ↓ ↓ ↓b 5 1 Note that studies counted as supporting may support only part of the results in a row, as not all parameters were studied in all counted references, and that the same paper may be counted as both supporting and contrasting when it supports the findings of some feeding parameters but contrasts those of others. Some papers included in the review were not included in this table due to incompatible study designs. a One result was not statistically tested. b This result was only seen in the morning. 5 J.D. Bus et al. Applied Animal Behaviour Science 241 (2021) 105383 show smaller drops in feed intake. The extent to which feeding duration size and meal duration, but an equal daily feeding duration, feeding reduces during clinical disease may be highly variable between pigs, frequency and proportion of feed eaten during the day, compared to pigs even when they are exposed to the same conditions. For example, pigs kept in hygienic conditions. More direct relationships between sub­ suffering from pneumonia reduced their daily feeding duration to be­ clinical disease and the feeding behaviours that underlie daily intake tween 0.33 and 48.33 min per day, maintained this for 1d to 12d, and have been reported as well. Maselyne et al. (2018) proposed that some completely stopped eating during 0d to 4d (Adrion et al., 2018). reductions in feeding duration, which could not be associated to clinical Moreover, reduced feeding duration may not occur during all types of disease, might have been due to subclinical disease, as post-mortem data disease, as it was reported for lame and coughing pigs but not in pigs showed that one such pig suffered from internal abscesses. In addition, suffering from diarrhoea or skin lesions (Kapun et al., 2016). This il­ pigs that were diagnosed with osteochondrosis after a CT scan, but lustrates the necessity of distinguishing between disease types when showed no lameness, visited the feeder more often, for a shorter dura­ studying impacts on feeding behaviours. It is currently unclear whether tion and at a higher feeding rate than pigs that did show lameness the reduction in daily feeding duration is the result of a reduced feeding (Munsterhjelm et al., 2017). Although these studies do not provide frequency or visit duration. One study reported that feeding frequency conclusive evidence as to how subclinical diseases impact feeding pat­ reduced after bacterial infection, but this was only seen in morning terns, they do illustrate that relations between feeding patterns and hours (Ahmed et al., 2014). In addition, Munsterhjelm et al. (2017) certain subclinical diseases exist. reported that pigs with clinical osteochondrosis had a lower feeding frequency and higher visit duration than pigs with subclinical osteo­ 4.3. Thermal stress chondrosis, but when compared to healthy control pigs, clinical osteo­ chondrosis solely led to a tendency for lower feeding frequency. It On the relation between feeding patterns and thermal stress, 27 should be noted, however, that in this study the control group was very papers were identified. Summaries of the main findings are provided in small (n = 3) and that a lack of statistical power may have obscured Table 3. differences between clinically ill and healthy pigs. Thermal stress occurs when ambient temperature is outside a pig’s The role of other feeding behaviours in relation to clinical disease is thermoneutral zone. In commercial husbandry, this can occur when currently unknown. Feeding rate was considered in only one study, outside temperatures are extreme or when climate-control technology which reported a lower rate in pigs with clinical compared to subclinical malfunctions. As the exact thermoneutral zone differs between studies, osteochondrosis, but a similar rate between clinically ill pigs and healthy depending on for example body weight and air humidity (National controls although the control group was very small (n = 3, Munsterhjelm Research Council, 2012), studies applying temperatures higher and et al., 2017). In addition, Kapun et al. (2017) proposed that a drop in the lower than the thermoneutral zone are accumulated as studies looking at number of pigs eating simultaneously may indicate lameness in the pig heat and cold stress, respectively. that became more solitary, or that lameness may be present in pigs that delay their first morning meal, but these potential indicators have not 4.3.1. Heat stress yet been tested empirically. Other parameters, such as visit interval or The effect of heat stress on pig feeding patterns is generally assessed the frequency of non-nutritive visits, have, to our knowledge, not been by increasing the temperature above the thermoneutral zone, either studied in relation to clinical disease. constantly or in a cyclic fashion, i.e. with higher temperatures during the day than night. Commonly, the impact of the high ambient temperature 4.2. Subclinical disease on the pigs is not further assessed, as only few studies have included measurements on body temperature (Lopez et al., 1991a; Morales et al., On the relation between feeding patterns and subclinical disease, 3 2018) or behavioural signs of heat stress, such as panting (Lopez et al., papers were identified. A summary of the main findings is provided in 1991a). Nevertheless, all studies have reported that pigs reduce their Table 2. feed intake during putative heat stress (Brown-Brandl et al., 2000; Collin Subclinical diseases are diseases that come with few to no clinical et al., 2001; dos Santos et al., 2018; Hyun et al., 1998b; Lopez et al., signs (i.e. they are asymptomatic) and are hence often detected only 1991a; Nienaber et al., 1987; Quiniou et al., 2000b, 2000a; Renaudeau after slaughter or when they proceed to become clinical. Although there et al., 2006; Rinaldo et al., 2000; Xin and DeShazer, 1992). This is an have been few direct studies into the relation between subclinical dis­ adaptive response, as feed intake during heat stress further increases ease and feeding patterns in pigs, there are some indications that sub­ body temperature, especially when intake occurs in the evening (Cer­ clinical diseases might alter feeding patterns. Although no studies have vantes et al., 2018). When ambient temperature returns to the thermo­ related subclinical diseases to daily feed intake, one study showed that neutral zone (Kerr et al., 2005) or pigs are sprayed with water pigs kept in low hygienic conditions, which predisposes pig to (sub) (Eigenberg et al., 2002), feed intake increases rapidly to levels similar to clinical disease, had a lower daily feed intake than pigs kept in clean or temporarily exceeding normal. When high temperatures persist, pigs conditions (Renaudeau, 2009). In addition, these authors found that pigs will over time increase their feed intake levels back to baseline (dos in unhygienic conditions changed some of their underlying feeding be­ Santos et al., 2018; Lopez et al., 1991a). The severity of the reduction in haviours: pigs in unhygienic conditions had a lower feeding rate, meal feed intake depends on pig factors, including breed (Cross et al., 2020; Table 2 The effect of subclinical disease on feeding patterns (intake (inta), duration (dur), frequency (freq) & rate) compared to feeding patterns of healthy pigs (increase ↑, decrease ↓, or no effect ≈). Empty cells show that no papers were found on this topic. The final column indicates the number of papers that support or contradict this finding. An overview of the papers this table is based on can be found in Supplementary Table 2. Daily level Meal level Visit level Reference count Subclinical disease Inta Dur Rate Inta Dur Freq Inta Dur Freq Support Contrast Low hygiene level ↓ ≈ ↓ ↓ ↓ ≈ ≈ 1 0 Internal abscesses ↓ 1a 0 Osteochondrosis ↓ ↓ ↑ 1 0 Note that studies counted as supporting may support only part of the results in a row, as not all parameters were studied in all counted references, and that the same paper may be counted as both supporting and contrasting when it supports the findings of some feeding parameters but contrasts those of others. a This result was not statistically tested, and is theorised from the findings of one pig only. 6 J.D. Bus et al. Applied Animal Behaviour Science 241 (2021) 105383 Table 3 The effect of high and low ambient temperatures on feeding patterns (intake (inta), duration (dur), frequency (freq) & rate) compared to feeding patterns of pigs at control conditions (treatment vs. control, increase ↑, decrease ↓, or no effect ≈). Empty cells show that no papers were found on this topic. In case of ties for the main effect, papers with animal-based parameters were given greater weight than those with resource-based parameters, and if this was insufficient both results are shown. The final column indicates the number of papers that support or contradict the main finding. An overview of the papers this table is based on can be found in Supplementary Table 3. Daily level Meal level Visit level Reference count Thermal stress Inta Dur Rate Inta Dur Freq Inta Dur Freq Support Contrast Constant hot (±30 C) vs. thermoneutral ◦ ↓ ↓ ↑≈ ↓ ↓ ≈ ↑ 6 2 Hot day vs. thermoneutral ↓ ≈ ≈ ≈ 3 0 Hot day & less hot night vs. constant hot ≈ ≈ 1 0 Large vs. small heat variation between day and night ↓ ≈ ≈ ↓ 2 0 Hot season or hot outside temperature vs. thermoneutral ↓a ↓b ≈ ↓ ↓ ↑ ≈ 4 3 Constant cold (1–13 ◦ C) vs. thermoneutral ↑ ↑ ↓ ↓ ↓≈ ↑ 5 0 Cold night & thermoneutral day vs. hot day & thermoneutral night ↑ ↑ ≈ ≈ ≈ 1 0 Note that studies counted as supporting may support only part of the results in a row, as not all parameters were studied in all counted references, and that the same paper may be counted as both supporting and contrasting when it supports the findings of some feeding parameters but contrasts those of others. Some papers included in the review were not included in this table due to incompatible study designs. a In one study, the results depended on age: ↑ from 15 to 30 kg, ↓ from 35 to 90 kg & overall from 15 to 90 kg. b In one study, the results depended on breed: ↑ for Yorkshire- and Duroc-sired pigs, ↓ for Landrace-sired pigs. Wellock et al., 2003) and age (Nienaber et al., 1996; Quiniou et al., as the Creole, or crosses of different breeds. Housing conditions ranged 2000a; Rinaldo et al., 2000), and environmental factors, including group from individual to group housing, and different diets and lighting re­ size (Kerr et al., 2005) and feeding regime (dos Santos et al., 2018). More gimes were used. It is hence conceivable that although the impacts of specifically: older and hence larger pigs are more affected by high heat stress on feed intake and duration are similar across circumstances, temperatures than younger and hence smaller pigs; individually-housed they result from divergent adaptations in the underlying feeding be­ pigs are more affected than group-housed pigs, but group-housed pigs haviours. These adaptions depend on, among others, pigs’ physical become more affected as pig densities increase; and precision feeding (i. abilities, their housing circumstances and the exact nature of the high e. where the exact ration is calculated for each pig individually) may ambient temperatures. reduce the impact of high temperatures because the optimised nutrient composition allows for a lower feed intake and hence less heat pro­ 4.3.2. Cold stress duction during digestion. In addition to reducing feed intake, pigs may Low ambient temperatures, in contrast to high ones, prompt pigs to shift their feeding (peaks) earlier into the morning and later into the increase their feed intake. The increased feed intake is required to night (Cross, 2017; Cross et al., 2020; dos Santos et al., 2018; Feddes maintain body temperature (Gertheiss et al., 2015; Lopez et al., 1991b; et al., 1989; Ingram et al., 1980; Quiniou et al., 2000a, 2000b; Renau­ Nienaber et al., 1990; Quiniou et al., 2000a; Whittemore et al., 2001b) deau et al., 2006; Xin and DeShazer, 1992), but not if the high tem­ and not for growth, as pigs show reduced growth during cold stress perature persists during the night (Collin et al., 2001). (Lopez et al., 1991b; Nienaber et al., 1987). Unlike heat stress, cold How these changes in feed intake are mediated by underlying stress does not induce pigs to eat at different day times than preferred feeding behaviours is not fully clear. There is general agreement that (Quiniou et al., 2000a). If cold stress persists, feed intake will reduce daily feeding duration declines as temperatures rise (Brown-Brandl back to baseline levels after about two weeks (Lopez et al., 1991b). The et al., 2000; Collin et al., 2001; Fraga et al., 2019; Gertheiss et al., 2015). extent to which feed intake increases depends, among others, on pig Although one study reported no such reduction (Renaudeau et al., body size and feed composition. Specifically, small pigs increase their 2006), in this observational study the difference in ambient temperature intake more than large pigs, indicating a higher susceptibility to cold between the warm and hot season was relatively small and may hence stress (Quiniou et al., 2000a). The extent to which intake can be have been insufficient to induce differences. Cross et al. (2018) deter­ increased appears constrained by gut capacity and feed type, as pigs fed mined that the extent of the reduction in feeding duration in response to on feed rich in fibres did not show this typical increase in intake heat stress is heritable, as it appears mediated by DNA sequences that (Whittemore et al., 2001b). It might, therefore, be that small pigs simply code for proteins involved in for example immune function, the endo­ cannot increase their intake as much as large pigs due to limitations in crine system and vasoconstriction. Lower daily feeding duration and gut capacity. intake may be due to reduced feeding frequency (Brown-Brandl et al., Only a few studies have looked at the impact of cold on the feeding 2000; Cross et al., 2020; Xin and DeShazer, 1992), feed intake per meal behaviours underlying intake, and their results are quite consistent. (Brown-Brandl et al., 2000; dos Santos et al., 2018) and meal duration Higher feed intake coincides with a longer daily feeding duration (Brown-Brandl et al., 2000; Nienaber et al., 1996; Renaudeau et al., (Gertheiss et al., 2015; Nienaber et al., 1990; Quiniou et al., 2000a), due 2006), but many studies also reported no impact of heat stress on these to a lower feeding rate (Nienaber et al., 1991, 1990), a higher feeding parameters (Collin et al., 2001; Nienaber et al., 1996; Quiniou et al., frequency (Nienaber et al., 1991, 1990), and a lower visit interval 2000a; Renaudeau et al., 2006; Xin and DeShazer, 1992). Feeding rate (Nienaber et al., 1991). Meanwhile, feed intake and feeding duration per has been reported to be higher (Brown-Brandl et al., 2000; Nienaber visit are reduced (Nienaber et al., 1991) or not affected (Quiniou et al., et al., 1996), lower (dos Santos et al., 2018) or unchanged (Collin et al., 2000a). Although these consistent results would suggest that the impact 2001; Feddes et al., 1989; Quiniou et al., 2000a; Renaudeau et al., 2006) of cold stress on pig feeding behaviour is more consistent than that of during high temperatures. The wide range of reported impacts of high heat stress, the number and diversity of studies on which this is based is temperatures on feeding behaviours is likely a consequence of the wide much smaller for cold than for heat stress and does hence not allow such range of treatments, animals and housing conditions used. Treatments a firm conclusion. Expanding the range of animals and cold treatments have consisted of cyclic or constant temperatures, with maximum tem­ used might reveal situation-specific adaptations to cold stress similar to peratures ranging from 25 to 40 ◦ C, or studies have simply followed those seen during heat stress. outside temperatures. Pig breeds used ranged from high-productive breeds such as Piétrain and Large White to heat-resistant breeds such 7 J.D. Bus et al. Applied Animal Behaviour Science 241 (2021) 105383 4.4. Social behaviour third theory suggests that it is the individually- rather than the group-housed pigs that show deviant feeding behaviour, as a lack of Pigs are a highly social species that naturally live in family groups stimuli induces them to visit the feeder more often. To our knowledge, and show elaborate social interactions (D’Eath and Turner, 2009). no attempts have yet been made to interpret empirical study results in Although housing with conspecifics is essential for the expression of light of the third theory. Below, we will discuss the evidence that un­ natural social behaviour, group housing also has an extensive impact on derlies the first two theories. In total, 58 papers were identified. A pig feeding patterns. Several modelling studies have addressed the summary of the main findings is provided in Table 4. importance of social interactions for feeding patterns (Boumans et al., 2018a, 2018b; Wellock et al., 2003). Genomic studies that modelled 4.4.1. Feed competition feeding behaviour have repeatedly shown that indirect genetic effects, Feed competition has been extensively studied in growing-finishing which largely reflect the social interactions between pen mates, pigs, due to its relationship with growth and carcass quality (i.e. contribute strongly to the variation in feeding behaviours between pigs lesioning). It mostly consists of one pig displacing another from the (Do et al., 2013a; Herrera-Cáceres et al., 2019; Lu et al., 2017; Ragab (individual) feeder, either with or without aggressive behaviours such as et al., 2019). This becomes even more pronounced when the indirect biting, and can in some cases lead to fights. Studies on feed competition genetic effects are calculated based on between-pig differences in generally focus on either resource- or animal-based parameters related feeding behaviours, to reflect that not all dyads of pigs have the same to competition. More specifically, studies either manipulate the physical level of interaction (i.e. pigs that have widely different feeding patterns or social environment to influence competition, e.g. by changing feeder are assumed to have a lower level of social interaction than pigs that had space or grouping pigs in high or low weight-diverse groups, or look at similar feeding patterns) (Ragab et al., 2019). In addition, remarkable the feeding patterns of pigs with different characteristics, such as differences in feeding patterns have been reported between pigs housed dominance ranks. individually and pigs housed in groups, where group-housed pigs could have either one individual feeder or sufficient feeder spaces to accom­ 4.4.1.1. Adaptations to resource-induced feed competition at pen level. modate all pigs simultaneously. Group-housed pigs show either a similar Feed competition can be manipulated using a wide range of resource (de Haer and de Vries, 1993) or a lower (de Haer et al., 1992; Gonyou changes, including space allowance, group size, feeder space, feeder et al., 1992; Kerr et al., 2005; Nielsen et al., 1996b) feed intake than pigs location, and how similarly sized the pigs in one pen are. In addition, housed individually, despite a higher intake per visit and meal (Bornett mixing pigs into new groups may temporarily increase feed competition, et al., 2000; de Haer and de Vries, 1993). These differences mainly result until a new dominance order has been established. In general, compe­ from a lower feeding frequency in group-housed than tition becomes worse when pen and feeder space decrease, group size individually-housed pigs (Bornett et al., 2000; de Haer et al., 1992; de increases, feeders are placed closer together, and when pigs become Haer and de Vries, 1993; Nielsen et al., 1996b), which is partly more similar in body weight, as pigs of similar weight have more trouble compensated by a higher visit or meal duration (Bornett et al., 2000; de establishing a clear dominance order (Andersen et al., 2000). The effect Haer and de Vries, 1993) and a higher feeding rate (Bornett et al., 2000; of these competition-influencing factors on feeding patterns has often de Haer et al., 1992; de Haer and de Vries, 1993). Group housing sup­ been studied in co-occurrence (e.g. group size was increased but the ports pigs’ diurnal, alternans feeding pattern, even when external Zeit­ number of feeders remained the same, confounding larger group size gebers are eliminated (Ingram et al., 1980). Currently, there are three with lower feeder allowance), making it difficult to pinpoint how each theories as to how group housing influences feeding patterns (Bornett factor independently influences feeding patterns. When the factors could et al., 2000; Nielsen, 1999). The first theory claims that pigs adapt their be split, they indicated similar effects on feeding patterns. Therefore, we feeding patterns in response to competition for access to the feed. The accumulated these factors as leading to high or low feed competition, second builds on the first, by suggesting that social facilitation (i.e. and present them as combined results (an overview of the split results observing another pig feeding induces a pig to start feeding itself) en­ can be found in Supplementary Table 4). hances feed competition and hence further alters feeding patterns. The In general, it has been reported that feed competition will induce Table 4 The effect of resource- and animal-based manipulations of social behaviour on pig feeding patterns (intake (inta), duration (dur), frequency (freq) & rate) (treatment vs. control, increase ↑, decrease ↓, or no effect ≈). Empty cells show that no papers were found on this topic. In case of ties for the main effect, papers with animal-based parameters were given greater weight than those with resource-based parameters, and if this was insufficient both results are shown. The final column indicates the number of papers that support or contradict the main finding. An overview of the papers this table is based on can be found in Supplementary Table 4. Daily level Meal level Visit level Reference count Social behaviour Inta Dur Rate Inta Dur Freq Inta Dur Freq Support Contrast Group vs. individual housing ↓ ↓≈ ↑ ↑ ↑ ↑ ↓ 8 3 More animals per feeder ≈ ↓ ↑ ↑ ↑↓ ↓ 12 8 Less trough space per pig ↓ ↓ ≈ ≈ 2 0 Feeders closer together ≈ ↓ 1 0 Absence vs. presence of protective crate ≈ ↓ ↑ 1 0 More pigs per pen ≈ ↓ ↑ ↓≈ ↑↓ ↑↓ ↑↓ ≈ ↓ 7a,b 3 Mixing vs. no mixing ≈↓ ≈ ≈ ↑ ↓ 2 0 Lower space allowance ≈↓ ≈ ≈ ↑ ↑↓≈ ↓ 8c 3 Dominant vs. subordinate (classification methods accumulated) ≈ ↑ ↑↓≈ ↑ ≈ ≈ ↑ ↑↓ ↑↓≈ 11 4 Large vs. small body size ↑ ↓ ↑≈ ≈ ↑ ↓ 3d 1 Note that studies counted as supporting may support only part of the results in a row, as not all parameters were studied in all counted references, and that the same paper may be counted as both supporting and contrasting when it supports the findings of some feeding parameters but contrasts those of others. Some papers included in the review were not included in this table due to incompatible study designs. a The effect of group size on feed intake per visit was inconsistent, with groups of 12 pigs having a lower intake per visit than groups of 8 pigs, while groups of 4 pigs had a lower intake per visit than groups of 12 or 8 pigs. b The effect of group size on feed intake per meal, meal duration and meal frequency differed between breeds. c The effect of lower space allowance on visit duration depended on age. d Medium-sized pigs showed a higher feeding rate than small or large pigs. 8 J.D. Bus et al. Applied Animal Behaviour Science 241 (2021) 105383 pigs to maintain daily feed intake by changing their underlying feeding distinct peaks (Botermans and Svendsen, 2000; Hyun and Ellis, 2002; pattern (Edwards et al., 1988; Gonyou and Lou, 2000; Hansen et al., Nielsen et al., 1995; Walker, 1991) and will shift part of their feeding 1982; Hyun et al., 1998a; Hyun and Ellis, 2002; McGlone and Newby, behaviour into the night (Botermans and Svendsen, 2000; Georgsson 1994; Mikesell and Kephart, 1999; Morrison et al., 2003; Nielsen et al., and Svendsen, 2002; Hyun and Ellis, 2002, 2001; Morrison et al., 2003; 1995; O’Connell et al., 2004; Pearce and Paterson, 1993; Rasmussen Morrow and Walker, 1994a; Nielsen et al., 1995; Turner et al., 2002; et al., 2006b; Schmolke et al., 2003; Spoolder et al., 1999; Tindsley and Walker, 1991). One study saw no effect on diurnal patterns (Simonsen, Lean, 1984; Wiegand et al., 1994). When these adaptations are insuffi­ 1990), but this study had only one group of pigs per treatment and hence cient, feed intake reduces (Edmonds et al., 1998; Georgsson and likely suffered from insufficient statistical power. Besides night feeding, Svendsen, 2001; Hyun et al., 1998a, 1998b; Hyun and Ellis, 2001; high feeder occupation also leads to an increase in queuing for access to McGlone and Newby, 1994; Morrow and Walker, 1994a; Turner et al., the feeder (Morrow and Walker, 1994a; Rasmussen et al., 2006a, 2006b; 2002; Walker, 1991). At pen level, pigs will reduce their daily feeding Spoolder et al., 1999; Walker, 1991), albeit not when a trough was used duration during competition (Botermans and Svendsen, 2000; Georgs­ (Turner et al., 2002). Finally, one study reported that, at pen level, the son and Svendsen, 2002; Gonyou and Lou, 2000; Hyun et al., 1998a; frequency of non-nutritive visits was not affected by competition Hyun and Ellis, 2001; Labroue et al., 1999; Morrow and Walker, 1994a; (Nielsen et al., 1995). Nielsen et al., 1995; Turner et al., 2002), and hence maintain their daily feed intake levels by increasing feeding rate (Botermans and Svendsen, 4.4.1.2. Animal-based adaptations to feed competition. Although the 2000; Hyun and Ellis, 2002, 2001; Labroue et al., 1999; Nielsen et al., previously described changes in feeding patterns due to different levels 1995; Rasmussen et al., 2006a). A few studies, however, reported no of feed competition occur on average, at the individual level pigs may effect of group size or space allowance on feeding duration (Hyun et al., show vastly different adaptations, especially when competition becomes 1998a; Hyun and Ellis, 2002; Pearce and Paterson, 1993; Spoolder et al., more severe (Chen et al., 2010; Walker, 1991). These differences in 1999) or rate (Hyun et al., 1998a), implying that feeder space is likely to adaptation strategy are largely due to differences in dominance rank, be the more determining factor. The changes in daily feeding duration which in turn is influenced by a range of pig characteristics, such as and rate are mediated by changes in the frequency and size of the visits breed, body weight, sex, and personality (Beilharz and Cox, 1967; and meals. Most studies have reported that, during competition, pigs Boumans et al., 2018a). Pigs may eat according to rank order only if reduce their feeding frequency (Georgsson and Svendsen, 2002; Hansen competition is present, eating randomly when competition is absent et al., 1982; Hyun et al., 1998a; Hyun and Ellis, 2002, 2001; Labroue (Hansen et al., 1982). Dominance orders can be determined in many et al., 1999; Morrison et al., 2003; Morrow and Walker, 1994a; Nielsen ways, such as by using agonistic interactions around the feeder (Hoy et al., 1995; Wallenbeck and Keeling, 2013) and increase their intake et al., 2012) or after mixing (Leiber-Schotte, 2009; Parois et al., 2017), (Georgsson and Svendsen, 2002; Hyun et al., 1998a; Hyun and Ellis, or by extrapolating feeding strategies from correlations between 2002; Nielsen et al., 1995) and duration (Hyun et al., 1998a; Hyun and different feeding behaviours (Herrera-Cáceres et al., 2019; Ragab et al., Ellis, 2002; Nielsen et al., 1995) per visit. Several studies have, however, 2019). Besides using dominance orders, many studies have also classi­ reported no differences (Gonyou and Lou, 2000; Labroue et al., 1999; fied pigs as dominant, intermediate or subordinate, using assumptions Morrison et al., 2003; Rasmussen et al., 2006b; Turner et al., 2002; based on body weight and gender: large pigs are thought to be more Walker, 1991), or even an increase in feeding frequency (Labroue et al., dominant than small pigs (Botermans and Svendsen, 2000; Georgsson 1999) and reduction in intake (Hyun and Ellis, 2001; Labroue et al., and Svendsen, 2002; Rasmussen et al., 2006b), and boars are thought to 1999) and duration (Hyun and Ellis, 2001; Labroue et al., 1999; Ras­ be most dominant, followed by barrows and finally gilts (Puppe et al., mussen et al., 2006b; Walker, 1991) per meal. It is likely that these 1991). All these methods have been applied to compare the feeding contrasting results are due to the specific study circumstances. For patterns between pigs of different ranks. Whether different methods example, the lack of changes in response to competition were likely result in the same dominance order is currently unclear, but could be because Labroue et al. (1999) used a different breed than all other questioned as, for example, agonistic interactions around the feeder may studies, and expressed feeding behaviour at meal rather than visit level. be partly driven by feeding motivation, unlike those after mixing. The deviating results of Turner et al. (2002) and Rasmussen et al. The differences in feed intake between pigs of different ranks are (2006b) were likely due to them using a trough rather than single-space inconsistent. When dominance order was inferred from pig body size, feeders, allowing multiple pigs to feed simultaneously. It has been the­ feed intake was reported to be higher in dominant pigs than in subor­ orised that the exact pattern shown depends on how easily pigs can be dinate pigs (Chen et al., 2010; Georgsson and Svendsen, 2002). How­ displaced from the feeder. If displacement is difficult, such as when a ever, when agonistic interactions were used to determine order, a lower protective crate is used (Morrow and Walker, 1994b), pigs are forced to feed intake in dominant pigs was observed (Hoy et al., 2012; Leiber-­ withdraw from the feeder less often but may have more difficulty in Schotte, 2009), or no difference (Nielsen et al., 1995). Therefore, it accessing the feeder, leading to a low feeding frequency and high seems likely that the higher feed intake is related to higher nutritional feeding duration per visit during competition. If displacement is easy, requirements of larger pigs, and not dominance rank per se. Neverthe­ pigs, especially small ones (Botermans et al., 2000), are forced to less, it has been reported that dominant pigs increase their feed intake withdraw often and will hence maintain their intake by visiting the when competition becomes more severe (Georgsson and Svendsen, feeder often but for small visits each time. This theory corresponds to the 2002), implying that, during competition, high-ranked pigs may domi­ results found in a modelling study by Boumans et al. (2018b), in which nate the feeder more than nutritionally required. the simulated pigs could be easily displaced, and from which the authors These changes in feed intake are underlain by large differences in the concluded that a low meal size and duration and a high meal frequency feeding patterns between dominant and subordinate pigs. Dominant indicate a competitive environment with high displacement levels. It pigs show a feeding behaviour referred to as ‘meal eating’, while sub­ hence appears that the exact environmental circumstances as well as ordinate pigs perform ‘nibbling’. More specifically, dominant pigs show pigs’ physical characteristics strongly influence how pigs adapt their a lower feeding frequency than subordinate pigs (Gonyou and Lou, feeding patterns to competition. 2000; Hoy et al., 2012; Ragab et al., 2019; Schamun and Hoy, 2011), Besides the behaviours underlying feed intake, several other feeding combined with a higher intake (Boumans et al., 2018a; Georgsson and parameters change in competitive situations and could hence potentially Svendsen, 2002; Hoy et al., 2012) and duration (Herrera-Cáceres et al., be used as indicators of pig welfare. Diurnal patterns in feeding become 2019; Hoy et al., 2012; Puppe et al., 1991; Ragab et al., 2019; Schamun less pronounced during competition: as feeder occupation increases and Hoy, 2011; Schönfelder, 2005) per visit or meal. These effects are (Hyun and Ellis, 2002, 2001; Morrison et al., 2003; Nielsen et al., 1995; likely due to subordinate pigs being the victim of forced withdrawals Walker, 1991), pigs will eat throughout the day rather than in two 9 J.D. Bus et al. Applied Animal Behaviour Science 241 (2021) 105383 more often than dominant pigs (Botermans and Svendsen, 2000). social facilitation has been frequently reported. Individually-housed Indeed, dominant pigs appear to gain access to the feeder more easily, as pigs were observed to eat simultaneously with a neighbouring pig they have a shorter waiting time than subordinate pigs (Rasmussen more often than could be expected by chance (Gonyou et al., 1992). In et al., 2006b), can feed more frequently during peak hours (Hoy et al., addition, individually-housed pigs have a higher feed intake when they 2012) and show a higher frequency of non-nutritive visits than subor­ are provided with neighbours rather than when kept in social isolation dinate pigs despite a lower daily feeding frequency (Georgsson and (Hsia and Wood-Gush, 1983), although this does not clarify whether it Svendsen, 2002). Contrasting results were reported by Leiber-Schotte was the feeding activity of these neighbours that induced the higher (2009), who found that dominant pigs had a lower feeding duration per intake. In a behavioural test, pigs were reported to have a higher feed day and visit than subordinate pigs but there was no difference in intake when the distance between feeders was smaller, despite having to feeding frequency, Nielsen et al. (1995), who found no correlations shift to another feeder more often (Thomsen et al., 2009). Interestingly, between dominance rank and any feeding behaviours, and Herrer­ pigs that were neighboured by a group of pigs on each side of the in­ a-Cáceres et al. (2019), who found that dominant pigs had more feeder dividual pen showed a similarly enhanced level of simultaneous feeding, visits than subordinate. These contrasts may have been due to the but only with one of the two neighbouring groups and only if the feeder different feeder types and methods used to determine the dominance in the group pen was immediately adjacent to the feeder in the indi­ order. Leiber-Schotte (2009) was the only one with a feeder with a full vidual pen (Gonyou et al., 1992). It has been reported that social facil­ protective crate. In addition, Leiber-Schotte (2009) used the outcome of itation can induce feeding behaviour even when a pig is not hungry, as fights after initial mixing (i.e. wins and losses) and Nielsen et al. (1995) pigs that ate to satiation after fasting were observed to resume feeding used all agonistic interactions, and were hence the only studies using a when a fasted pig began eating in an adjacent pen (Hsia and Wood-Gush, scoring method uninfluenced by feeding motivation. Herrera-Cáceres 1984b). Interestingly, resumed feeding occurred more extensively in et al. (2019) used phenotypic and genetic correlations to identify dominant than submissive pigs, however, it should be noted that pigs behavioural strategies, and solely saw that pigs that occupied the feeder were classified as dominant using competition at the feeder and may longer and ate at a low rate, which were assumed to be dominant pigs, hence also have been simply more motivated to feed. Although these also had more feeder visits with shorter visit intervals. These correla­ studies show that individually-housed pigs, with unlimited access to a tions combined with indirect genetic effects also suggested that domi­ feeder, are influenced by social facilitation, these results cannot be nant pigs carry genes that induce them to stay longer at the feeder, while extrapolated to group-housed pigs, which are subject to competition for simultaneously carrying genes that force their pen mates into a subor­ access to the feeder. dinate role (Herrera-Cáceres et al., 2019). Subordinate pigs then have a Studies in group-housed pigs have shown less consistent results than lower feeding frequency due to difficulty in accessing the feeder (Her­ studies in individually-housed pigs regarding the occurrence of social rera-Cáceres et al., 2019), or, as was reported in other studies, try to facilitation. Gonyou et al. (1992) reported that group-housed pigs fed adapt to frequent displacement by increasing feeding frequency, from a two-hole feeder ate simultaneously less often than could be ex­ reducing the size of each visit, and shifting their feeding behaviour away pected by chance, suggesting that observing another pig feed actually from peak hours, i.e. into the night (Botermans and Svendsen, 2000; deterred pigs from feeding, for example due to competition. These re­ Georgsson and Svendsen, 2002). sults, however, could also be explained by a preference for a certain Whether pigs also individually adapt by changing their feeding rate feeding space of the two-hole feeder and a willingness to wait for its is not entirely clear, as the results are contradicting. Although most availability, rather than avoidance of simultaneous feeding per se. studies reported a higher feeding rate in subordinate than in dominant Regarding feed intake, pigs kept in a group, but provided with only one pigs (Herrera-Cáceres et al., 2019; Ragab et al., 2019), one study re­ feeding space, showed higher intake levels than pigs housed in social ported a higher feeding rate in dominant pigs (Chen et al., 2010), which isolation, but similar to those of pigs housed individually with neigh­ was possibly due to subordinate pigs eating at a low feeding rate outside bours (Hsia and Wood-Gush, 1983). When the feeder space was of peak hours (i.e. at night). It has also been theorised that adaptation increased by provision of a long feeding trough, feed intake increased through feeding rate is limited by pigs’ physical abilities: in a again. These results suggest that social facilitation indeed increases feed competition-free testing environment, Gonyou and Lou (2000) saw that intake, but that the extent of this increase is limited by feed competition. large pigs have a higher feeding rate than small pigs. Moreover, Alternatively, the pigs kept in social isolation may have had a reduced Georgsson and Svendsen (2002) proposed that small and large pigs show intake due to a lack of stimulation. With these contrasting results, it a similar feeding rate when kept in the same pen because small pigs do could be questioned whether social facilitation is an important not have the oral capacity to increase their feeding rate any further in contributor to the feeding behaviour of group-housed pigs. Indeed, a response to competition. Instead, pigs of intermediate size, which were recent modelling study concluded that social facilitation only has a assumed to have intermediate rank, showed the highest feeding rate, as minor impact on pig feeding patterns, as adding social facilitation to the these were physically capable of increasing their feeding rate in response model had little effect on the simulated feeding behaviours (Boumans to competition. Two modelling studies support the effect of body weight et al., 2018b). in adaptation to competition: Wellock et al. (2003) reported that pigs with high growth potential are more severely impacted by space 4.5. Tail biting allowance restrictions than pigs with low growth potential, as they reduced feed intake more severely, and Boumans et al. (2018a) reported The most well-known abnormal behaviour in commercial pigs is tail that almost all feeding behaviours were influenced by growth capacity. biting. During tail biting, pigs nibble or bite the tail of a pen mate. This, In addition, the latter authors proposed that pig coping style or preferred in its severest form, causes wounds or even loss of the victim’s tail. In the behavioural strategy to deal with conflict might interact with domi­ biter, the behaviour is thought to be indicative of inadequate coping nance in determining individual feeding patterns during competition with the environment, and hence the performance of tail biting is, in (Boumans et al., 2018b, 2018a), however, this has, to our knowledge, itself, interpreted as reduced welfare. No papers could be identified that not been empirically assessed. looked at the feeding patterns of biters, but a few papers have reported the feeding patterns of tail-biting victims. Generally, a tail-biting victim 4.4.2. Social facilitation is defined as a pig with a fresh tail wound, and the onset of a tail biting It has been theorised that watching other pigs eat will induce feeding outbreak is defined as the moment at which at least one pig in the pen behaviour in the observing pig, a theory referred to as social facilitation. shows a fresh tail wound. In total, 4 papers were identified. A summary As social facilitation would lead to more pigs trying to eat simulta­ of the main findings is provided in Table 5. neously, it could enhance feed competition. In individually-housed pigs, At pen level, pigs in pens with at least one tail biting victim showed a 10 J.D. Bus et al. Applied Animal Behaviour Science 241 (2021) 105383 Table 5 The effect of tail wounds on feeding patterns (intake (inta), duration (dur), frequency (freq) & rate) compared to feeding patterns of pigs with intact tails (increase ↑, decrease ↓, or no effect ≈). Empty cells show that no papers were found on this topic. The final column indicates the number of papers that support or contradict the main finding. An overview of the papers this table is based on can be found in Supplementary Table 5. Daily level Meal level Visit level Reference count Tail wounds Inta Dur Rate Inta Dur Freq Inta Dur Freq Support Contrast Tail wounds (pen level) ↓ 1 0 Tail wounds (individual level) ↓ ↓↑ 1a 0 Some papers included in the review were not included in this table due to incompatible study designs. a Effects depended on age. lower feeding frequency 6− 9wks before outbreak onset (Wallenbeck 4.6. Environmental enrichment and Keeling, 2013) and on the day the first tail wound was observed (Viitasaari et al., 2015). Within outbreak pens, however, the victim was Recent frameworks have expanded the concept of animal welfare reported to have a higher feeding frequency 2− 5wks before outbreak from a focus on purely the prevention of negative states (including both onset compared to non-bitten pen mates or pigs in pens with no tail physiological and mental states) to the inclusion of positive states biting (Wallenbeck and Keeling, 2013). During the last few weeks before (Boissy and Erhard, 2014; Green and Mellor, 2011; Yeates, 2011). How outbreak onset, pigs in tail-biting pens had a lower feed intake than pigs positive states can be identified exactly is still in development, but ex­ in control pens (Munsterhjelm et al., 2015; Wallenbeck and Keeling, amples of proposed indicators include cognitive, such as judgement bias, 2013). The severity of this reduction is lower than the reduction seen in behavioural, such as play or certain vocalisations, and physiological lame pigs and has a faster recovery (Munsterhjelm et al., 2015) but, indicators, such as immunological or neurological markers (Yeates and nevertheless, reduced intake can persist for up to two weeks after Main, 2008). No studies have yet attempted to relate animal-based in­ outbreak onset (Wallenbeck and Keeling, 2013). The severity of intake dicators of positive states to pig feeding patterns. Nevertheless, several reduction was reported to be larger in younger pigs and in pigs that do studies have looked at the impact of providing pigs with environmental not recover (Munsterhjelm et al., 2015). Although the reduction in feed enrichment, mostly in the form of rooting substrates, on pig feeding intake and feeding frequency once a tail wound has emerged can likely patterns, which could be theorised to stimulate positive states. In total, be related to pain and stress, it could be questioned whether the changes 13 papers were identified. A summary of the main findings is provided in in feeding behaviours prior to outbreak onset were the result of early Table 6. stages of tail biting and social stress, which had yet gone unnoticed, or Both straw-based deep litter housing (Morrison et al., 2007) and whether these behaviours rather predisposed pigs to tail biting. For simple addition of straw to the pen (Bolhuis et al., 2006; Jensen et al., example, pigs with a higher feeding frequency may have to queue for the 2019) as environmental enrichment have been reported to increase feed feeder more often, making them an easy target for tail biting (Wallen­ intake. However, in one study this increase was only seen in the morning beck and Keeling, 2013). Indeed, when pigs were given a social rank and disappeared when pigs became older (Fraser et al., 1991), and in based on their feeding frequency and feed intake per visit (i.e. another straw availability only increased intake if straw had also been low-ranked pigs were assumed to have a high feeding frequency and a available during the rearing phase (Bolhuis et al., 2006). It might be that low intake per visit), it was seen that low-ranked pigs had more tail straw increases feed intake mostly at a young age, as Peeters et al. lesions at the beginning of the growing period than high-ranked pigs, but (2006) reported that straw does not affect feed intake when provided this relationship disappeared before slaughter (Heckmann et al., 2018). only during the last six weeks before slaughter. One study reported no These results suggest that pigs with a higher feeding frequency may be difference in feed intake in deep litter-housed pigs (Wei et al., 2019), more likely to become a victim of tail biting. Regardless of whether the however, pigs in this study did not have ad libitum access to feed and altered feeding behaviour is a cause or result of tail biting, the results were hence likely limited in increasing their intake by competition and imply that tail-biting outbreaks may be detected from a low average physiological constraints. Although the reported increases in feed intake feeding frequency at pen level and a reduction in feed intake, while the could be related to more improved welfare, there could be several other future victim could possibly be identified from its high feeding explanations. For example, higher intake may be due to physiological frequency. changes, such as better gut health (Jensen et al., 2019). Indeed, other No studies could be identified that related feeding patterns to other forms of environmental enrichment may have very different effects on types of abnormal behaviours, such as ear biting, mounting, or belly feeding behaviour. Providing pigs with mushroom compost as enrich­ nosing (Brunberg et al., 2011). ment led to a lower rather than a higher feed intake compared to pigs not provided with any enrichment, although intake levels were not lower than in pigs provided with an empty rack (Beattie et al., 2001). In Table 6 The effect of environmental enrichment on feeding patterns (intake (inta), duration (dur), frequency (freq) & rate) compared to feeding patterns of pigs without enrichment (increase ↑, decrease ↓, or no effect ≈). Empty cells show that no papers were found on this topic. In case of ties for the main effect, papers with animal- based parameters were given greater weight than those with resource-based parameters, and if this was insufficient both results are shown. The final column indicates the number of papers that support or contradict the main finding. An overview of the papers this table is based on can be found in Supplementary Table 6. Daily level Meal level Visit level Reference count Environmental enrichment Inta Dur Rate Inta Dur Freq Inta Dur Freq Support Contrast Deep litter housing ↑≈ ↑ ↑ ↓ 3 0 Availability of straw ≈ ≈ ≈ ≈ ≈ ↑≈ 5 3 Availability of other rooting material ↓≈ ↓ 2 0 Rotation of toys ↓ 1 0 Group-farrowing with creep feed in early life ↓ 1 0 Note that studies counted as supporting may support only part of the results in a row, as not all parameters were studied in all counted references, and that the same paper may be counted as both supporting and contrasting when it supports the findings of some feeding parameters but contrasts those of others. 11 J.D. Bus et al. Applied Animal Behaviour Science 241 (2021) 105383 addition, exposing piglets to group-farrowing during the first 8 weeks of time scale, within the day. Currently, most studies have been performed life, which is thought to be better for welfare than using daily averages, but there are indications that some feeding be­ confinement-farrowing, decreased feed intake in the first 6 weeks after haviours may only deviate during parts of the day, often in the morning arriving in the growing-finishing unit (Li et al., 2012). (Ahmed et al., 2014; Fraser et al., 1991; Kapun et al., 2017). In addition, Very little is known about how the behaviours underlying feed intake studies have often been performed at either pen or pig level, while it may are affected by environmental enrichment. One study showed that pigs be advantageous to combine both pen and pig level observations, such as in deep litter housing at a higher space allowance had a lower feeding was reported for tail biting outbreaks (Wallenbeck and Keeling, 2013). frequency and higher meal duration than pigs housed on slatted floors at These paths warrant further research attention. a lower space allowance (Tallet et al., 2013). Another study, however, On the behavioural side, mostly the feeding behaviours that underlie reported the opposite pattern, where pigs provided with straw had a intake, rather than feed intake itself, deviate from normal during higher feeding frequency and a lower intake per visit, leading to a stressful situations. Feed competition appears to have a strong effect on shorter daily feeding duration, than pigs denied straw (Morgan et al., underlying feeding behaviours, but not intake, influencing both the 1998). These authors also reported that pigs provided with straw had a average pattern of the group and the variation in patterns between pigs clearer alternans feeding pattern with higher and earlier feeding peaks within a pen. Nevertheless, it could be questioned whether these ad­ than pigs without straw. In a follow-up study, however, none of these aptations should be considered indicative of reduced welfare, as they effects could be reproduced (Morgan et al., 1999), to which the authors might be a reasonable coping mechanism with which pigs adapt effec­ concluded that feeding frequency was only higher after pigs switched tively to their environment. It is currently unclear whether severe social from no straw to straw, and became lower when pigs were switched stress induces different feeding patterns that can be distinguished from from straw to no straw. It is likely that the contrasting results of these ‘acceptable’ variation due to effective coping, although it could be studies are due to specific housing circumstances. For example, Morgan theorised that this point is reached when daily feed intake reduces. For et al. (1999, 1998) theorised that the higher feeding frequency was due abnormal behaviours, the presence of stress is clearer, and as mentioned to the straw rack being closer to the feeder than the resting area, giving previously, for tail biting it appears possible to identify tail biting vic­ pigs with straw a lower cost of feeding than pigs without straw, as they tims based on pig and pen level deviations in feeding patterns. Whether already spent more time close to the feeder. In Tallet et al. (2013), on the these results extrapolate to other forms of abnormal behaviour, such as other hand, straw was provided throughout the pen, hence no such cost ear biting, mounting or belly nosing, is currently unknown. In addition, reduction was present. Two more studies reported no effect of straw it is unknown whether biters, in addition to victims, also show de­ bedding (Peeters et al., 2006) or wood chip provision (Jensen and viations in their feeding patterns, which is highly relevant from a Pedersen, 2010) on feeding duration. It should be noted that both of practical perspective as it would allow the farmer to identify and remove these studies had a very limited observation period (once per week 5 the biter from the pen rather than its victims. min, or 60 min after feeding) and hence are unlikely to provide a good Although there is currently a large number of studies that consider representation of daily feeding duration. Finally, one study reported a physiological and behavioural aspects of welfare, the affective part of decrease in daily feeding duration when pigs were provided with a welfare has been largely overlooked in the context of pig feeding pat­ rotation of toys (Pearce and Paterson, 1993), but this was only tested terns. Although negative affective states, such as pain or frustration, under overcrowded conditions. could reasonably be assumed for the different physiological and behavioural welfare problems covered in this review, explicit links be­ 5. Discussion tween specific affective states and these welfare problems have not yet been made. It is, for example, yet unknown whether other negative af­ The aim of this study was to review the current knowledge on the fective states, such as depression, are linked to specific feeding patterns, relationships between feeding patterns and the welfare of growing- and on the positive side of affect only very little knowledge is available. finishing pigs. In the literature, numerous links between feeding pat­ Although some studies have looked at the effects of environmental terns and pig welfare have been described, implying that data from enrichment, mostly rooting material, on feeding patterns, their effects electronic feeding stations is indeed a promising tool to continuously are inconsistent. It seems likely that these effects depend on the exact monitor pig welfare. Deviations from ‘normal’ feeding behaviour may enrichment and environment provided, and hence may not be a good occur swiftly in response to reduced welfare, and may sometimes pre­ representation of positive affect. Based on knowledge of resilience in cede other clinical or behavioural manifestations of reduced welfare, animals (Colditz and Hine, 2016), it could be theorised that pigs with such as for clinical disease and tail biting. These links are particularly positive affective states may have feeding patterns that are stable over established between feed intake and physiological welfare problems, time. Animal-based measurements of both positive and negative affec­ such as clinical health, thermal stress and tail biting wounds. The tive states are required to gain more knowledge on the potential re­ severity of reduction in intake in response to these physiological prob­ lationships between affective state and pig feeding patterns. lems could even enable distinguishing clinical disease from tail biting, as We want to emphasize that many studies have reported a large with clinical disease the reductions seem more severe. The behaviours variation in feeding patterns at least between pigs, but potentially also underlying feed intake provide further information, as they showed within pigs over time. This implies that it might be difficult to accurately distinct deviations for different disease types and may allow early isolate the variation in feeding patterns that is informative of pig wel­ detection of disease at a subclinical stage. Combining feed intake with its fare. Therefore, it is important to further understand this variation, underlying behaviours (feeding frequency, duration and rate), therefore, especially at the level of the individual pig. It is conceivable that pigs may be more effective in detecting and differentiating reduced welfare with different basal feeding patterns, for example due to differences in states. A better understanding of how these underlying feeding behav­ dominance rank, may respond to similar welfare problems in a different iours respond to a wider range of clinical and subclinical diseases is, way. This finding suggests that a combination of group and individual however, warranted before this knowledge can be utilised as a welfare feeding patterns should be used to monitor pig welfare. Individual assessment tool. In addition, it should be noted that the responses of feeding stations are ideal to obtain this group and individual feeding these underlying behaviours may in some cases be too diverse between pattern data, and additionally create opportunities to combine feeding different housing conditions to be good, generic indicators of welfare. patterns with other sensors that provide additional information on pig This is the case, for example, with thermal stress, where changes in production and welfare. As each pig that enters the feeder will be iso­ feeding patterns are strongly dependent on pig characteristics, housing lated, individually recognised, and will stand still within the feeder, data and management. An interesting finding is that, in some cases, there could for example be gathered on body weight (with a scale or 3D may be additional value in studying the feeding patterns on a shorter cameras), body temperature (with thermal imaging or an IR 12 J.D. Bus et al. Applied Animal Behaviour Science 241 (2021) 105383 thermometer) or even facial expressions (with a (3D) camera). Beattie, V.E., Sneddon, I.A., Walker, N., Weatherup, R.N., 2001. Environmental enrichment of intensive pig housing using spent mushroom compost. Anim. Sci. 72, 35–42. https://doi.org/10.1017/S1357729800055533. 6. Conclusion Beilharz, R.G., Cox, D.F., 1967. Social dominance in swine. Anim. Behav. 15, 117–122. https://doi.org/10.1016/S0003-3472(67)80021-6. Berckmans, D., 2014. Precision livestock farming technologies for welfare management Numerous relationships between feeding patterns and growing- in intensive livestock systems. Sci. Tech. Rev. Off. Int. Epizoot. 33, 189–196. https:// finishing pig welfare were identified. Reduced welfare states are doi.org/10.20506/rst.33.1.2273. mostly associated with deviations from basal feeding patterns. Physio­ Bergstrom, J.R., Nelssen, J.L., Edwards, L.N., Tokach, M.D., Dritz, S.S., Goodband, R.D., Derouchey, J.M., 2012. Effects of feeder design and changing source of water to a logical welfare problems strongly affect feed intake, and different types location separate from the wet-dry feeder at 4 or 8 weeks before harvest on growth, of physiological problems may induce varying deviations in the under­ feeding behavior, and carcass characteristics of finishing pigs. J. Anim. Sci. 90, lying feeding behaviours. Behavioural welfare problems are mostly re­ 4567–4575. https://doi.org/10.2527/jas.2011-4486. flected by changes in the underlying feeding behaviours, while feed Bigelow, J.A., Houpt, T.R., 1988. Feeding and drinking patterns in young pigs. Physiol. Behav. 43, 99–109. https://doi.org/10.1016/0031-9384(88)90104-7. intake remains relatively constant. This knowledge could be used to Boissy, A., Erhard, H.W., 2014. How studying interactions between animal emotions, develop algorithms that can automatically relate pig feeding patterns to cognition, and personality can contribute to improve farm animal welfare. In: pig welfare, at the individual level. It appears promising to use several Grandin, T., Deesing, M.J. (Eds.), Genetics and the Behavior of Domestic Animals. Academic Press, pp. 81–113. https://doi.org/10.1016/B978-0-12-394586-0.00003- feeding behaviours simultaneously, using both group- and individual- 2. level feeding patterns, at a short time scale (i.e. within a day). The Bolhuis, J.E., Schouten, W.G.P., Schrama, J.W., Wiegant, V.M., 2006. Effects of rearing outcomes of these algorithms could provide continuous rather than and housing environment on behaviour and performance of pigs with different coping characteristics. Appl. Anim. Behav. Sci. 101, 68–85. https://doi.org/ sporadic information on generic welfare, which could even be assessed 10.1016/j.applanim.2006.01.001. and summarised retrospectively (i.e. after slaughter). Bornett, H.L.I., Morgan, C.A., Lawrence, A.B., Mann, J., 2000. The effect of group Nevertheless, it should be considered that a large variation in feeding housing on feeding patterns and social behaviour of previously individually housed growing pigs. Appl. Anim. Behav. Sci. 70, 127–141. https://doi.org/10.1016/S0168- behaviour is present between and potentially within pigs, and this 1591(00)00146-5. variation should be well-understood before the variation that represents Botermans, J.A.M., Svendsen, J., 2000. Effect of feeding environment on performance, pig welfare can be interpreted. In addition, to get a good representation injuries and behaviour in growing-finishing pigs: group-based studies. Acta Agric. Scand. A - Anim. Sci. 50, 237–249. https://doi.org/10.1080/090647000750069430. of pig welfare and to be able to distinguish between different causes of Botermans, J.A.M., Georgsson, L., Weström, B.R., Olsson, A.C., Svendsen, J., 2000. Effect reduced welfare, more knowledge is required on the responses of the of feeding environment on performance, injuries, plasma cortisol and behaviour in underlying feeding behaviours to different welfare problems, the effects growing-finishing pigs: studies on individual pigs housed in groups. Acta Agric. of performing abnormal behaviours on feeding patterns and on how the Scand. A Anim. Sci. 50, 250–262. https://doi.org/10.1080/090647000750069449. Boumans, I.J.M.M., Bokkers, E.A.M., Hofstede, G.J., de Boer, I.J.M., 2015. Understanding affective aspect of welfare relates to pig feeding patterns. feeding patterns in growing pigs by modelling growth and motivation. Appl. Anim. Behav. Sci. 171, 69–80. https://doi.org/10.1016/j.applanim.2015.08.013. Boumans, I.J.M.M., de Boer, I.J.M., Hofstede, G.J., Bokkers, E.A.M., 2018a. Unravelling Declaration of Competing Interest variation in feeding, social interaction and growth patterns among pigs using an agent-based model. Physiol. Behav. 191, 100–115. https://doi.org/10.1016/j. physbeh.2018.03.030. The authors report no declarations of interest. Boumans, I.J.M.M., de Boer, I.J.M., Hofstede, G.J., Bokkers, E.A.M., 2018b. How social factors and behavioural strategies affect feeding and social interaction patterns in Acknowledgements pigs. Physiol. Behav. 194, 23–40. https://doi.org/10.1016/j.physbeh.2018.04.032. Broom, D.M., 2010. Animal welfare: an aspect of care, sustainability, and food quality required by the public. J. Vet. Med. Educ. 37, 83–88. https://doi.org/10.3138/ This research is part of the project ‘ClearFarm’, supported by the jvme.37.1.83. European Union’s Horizon 2020 research and innovation programme Brown-Brandl, T., Nienaber, J., Turner, L., Yen, J., 2000. Manual and thermal induced feed intake restriction on finishing barrows. I: effects on growth, carcass [grant number 862919]. composition, and feeding behavior. Trans. Am. Soc. Agric. Eng. 43, 987–992. Brown-Brandl, T., Rohrer, G.A., Eigenberg, R.A., 2013. Analysis of feeding behavior of Appendix A. Supplementary data group housed growing–finishing pigs. Comput. Electron. Agric. 96, 246–252. https://doi.org/10.1016/j.compag.2013.06.002. Brown-Brandl, T.M., Jones, D.D., Eigenberg, R.A., 2016. Modeling feeding behavior of Supplementary material related to this article can be found, in the swine to detect illness. In: CIGR-AgEng Conference. Aarhus, Denmark, pp. 1–6. online version, at doi:https://doi.org/10.1016/j.applanim.2021.10 Brunberg, E., Wallenbeck, A., Keeling, L.J., 2011. Tail biting in fattening pigs: associations between frequency of tail biting and other abnormal behaviours. Appl. 5383. Anim. Behav. Sci. 133, 18–25. https://doi.org/10.1016/j.applanim.2011.04.019. Carcò, G., Dalla Bona, M., Carraro, L., Latorre, M.A., Fondevila, M., Gallo, L., References Schiavon, S., 2018. Influence of mild feed restriction and mild reduction in dietary amino acid content on feeding behaviour of group-housed growing pigs. Appl. Anim. Behav. Sci. 198, 27–35. https://doi.org/10.1016/j.applanim.2017.09.020. Adrion, F., Brown-Brandl, T.M., Jones, D., Gallmann, E., 2018. Identifying anomalous Casey, D.S., 2003. Performance of growing and finishing boars and gilts fed using decreases in feeding time of grow-finish pigs. In: 10th International Livestock electronic versus conventional feeders. The Use of Electronic Feeders in Genetic Environment Symposium (ILES X). Omaha, Nebraska, USA. https://doi.org/ Improvement Programs for Swine. Iowa State University Capstones, pp. 93–116. 10.13031/iles.18-042. Castilha, L.D., Sangali, C.P., Esteves, L.A.C., Muniz, C.F., Furlan, A.C., Vasconcellos, R.S., Ahmed, S.T., Mun, H.S., Yoe, H., Yang, C.J., 2014. Monitoring of behavior using a video- Pozza, P.C., 2016. Day-night behaviour and performance of barrows and gilts (70- recording system for recognition of Salmonella infection in experimentally infected 100 kg) fed low protein diets with different levels of tryptophan and B6 vitamin. growing pigs. Animal 9, 115–121. https://doi.org/10.1017/S1751731114002213. Appl. Anim. Behav. Sci. 180, 35–42. https://doi.org/10.1016/j. Andersen, I.L., Andenæs, H., Bøe, K.E., Jensen, P., Bakken, M., 2000. The effects of applanim.2016.03.017. weight asymmetry and resource distribution on aggression in groups of CBS, 2015. Factsheet Varkens. unacquainted pigs. Appl. Anim. Behav. Sci. 68, 107–120. https://doi.org/10.1016/ Cervantes, M., Antoine, D., Valle, J.A., Vásquez, N., Camacho, R.L., Bernal, H., S0168-1591(00)00092-7. Morales, A., 2018. Effect of feed intake level on the body temperature of pigs Andretta, I., Pomar, C., Kipper, M., Hauschild, L., Rivest, J., 2016. Feeding behavior of exposed to heat stress conditions. J. Therm. Biol. 76, 1–7. https://doi.org/10.1016/j. growing–finishing pigs reared under precision feeding strategies. J. Anim. Sci. 94, jtherbio.2018.06.010. 3042–3050. https://doi.org/10.2527/jas.2016-0392. Chemineau, P., 2016. Invited review: importance of animal health and welfare for the Aschoff, J., 1966. Circadian activity pattern with two peaks. Ecology 47, 657–662. stability of the three pillars of sustainability of livestock systems. Adv. Anim. Biosci. Augspurger, N.R., Ellis, M., 2002. Weighing affects short-term feeding patterns of 7, 208–214. https://doi.org/10.1017/s2040470016000145. growing-finishing pigs. Can. J. Anim. Sci. 82, 445–448. https://doi.org/10.4141/ Chen, C.Y., Misztal, I., Tsuruta, S., Herring, W.O., Holl, J., Culbertson, M., 2010. A01-046. Influence of heritable social status on daily gain and feeding pattern in pigs. J. Anim. Augspurger, N.R., Ellis, M., Hamilton, D.N., Wolter, B.F., Beverly, J.L., Wilson, E.R., Breed. Genet. 127, 107–112. https://doi.org/10.1111/j.1439-0388.2009.00828.x. 2002. The effect of sire line on the feeding patterns of grow-finish pigs. Appl. Anim. Colditz, I.G., Hine, B.C., 2016. Resilience in farm animals: biology, management, Behav. Sci. 75, 103–114. https://doi.org/10.1016/S0168-1591(01)00188-5. breeding and implications for animal welfare. Anim. Prod. Sci. 56, 1961–1983. Baumung, R., Lercher, G., Willam, A., Sölkner, J., 2006. Feed intake behaviour of https://doi.org/10.1071/AN15297. different pig breeds during performance testing on station. Arch. Anim. Breed. 49, 77–88. https://doi.org/10.5194/aab-49-77-2006. 13 J.D. Bus et al. Applied Animal Behaviour Science 241 (2021) 105383 Collin, A., van Milgen, J., Dubois, S., Noblet, J., 2001. Effect of high temperature on lines of growing-finishing pigs under daily cyclic high ambient temperature feeding behaviour and heat production in group-housed young pigs. Br. J. Nutr. 86, conditions. J. Anim. Sci. 97, 2493–2504. https://doi.org/10.1093/jas/skz123. 63–70. https://doi.org/10.1079/bjn2001356. Frank, J.W., Mellencamp, M.A., Carroll, J.A., Boyd, R.D., Allee, G.L., 2005. Acute feed Cross, A.J., 2017. Genetic Factors Associated with Thermal Tolerance in Grow-finish Pigs intake and acute-phase protein responses following a lipopolysaccharide challenge as Measured by Feeding Behavior. South Dakota State University. in pigs from two dam lines. Vet. Immunol. Immunopathol. 107, 179–187. https:// Cross, A.J., Keel, B.N., Brown-Brandl, T.M., Cassady, J.P., Rohrer, G.A., 2018. Genome- doi.org/10.1016/j.vetimm.2005.04.014. wide association of changes in swine feeding behaviour due to heat stress. Genet. Sel. Fraser, D., Phillips, P.A., Thompson, B.K., Tennessen, T., 1991. Effect of straw on the Evol. 50, 1–12. https://doi.org/10.1186/s12711-018-0382-1. behaviour of growing pigs. Appl. Anim. Behav. Sci. 30, 307–318. https://doi.org/ Cross, A.J., Brown-Brandl, T.M., Keel, B.N., Cassady, J.P., Rohrer, G.A., 2020. Feeding 10.1016/0168-1591(91)90135-K. behavior of grow-finish swine and the impacts of heat stress. Transl. Anim. Sci. 4, Garrido-Izard, M., Correa, E.C., Requejo, J.M., Diezma, B., 2020. Continuous monitoring 1–7. https://doi.org/10.1093/tas/txaa023. of pigs in fattening using a multi-sensor system: behavior patterns. Animals 10, 17. Czycholl, I., Kniese, C., Schrader, L., Krieter, J., 2017. Assessment of the multi-criteria https://doi.org/10.3390/ani10010052. evaluation system of the Welfare Quality® protocol for growing pigs. Animal 11, Georgsson, L., Svendsen, J., 2001. One or two feeders for groups of 16 growing-finishing 1573–1580. https://doi.org/10.1017/S1751731117000210. pigs: effects on health and production. Acta Agric. Scand. - Sect. A Anim. Sci. 51, D’Eath, R.B., Turner, S.P., 2009. The natural behaviour of the pig. In: Marchant-Forde, J. 257–264. https://doi.org/10.1080/09064700152717245. (Ed.), The Welfare of Pigs. Springer, Dordrecht, Netherlands, pp. 13–45. https://doi. Georgsson, L., Svendsen, J., 2002. Degree of competition at feeding differentially affects org/10.1111/j.1464-410X.2006.05867.x. behavior and performance of group-housed growing-finishing pigs of different D’Eath, R.B., Jack, M., Futro, A., Talbot, D., Zhu, Q., Barclay, D., Baxter, E.M., 2018. relative weights. J. Anim. Sci. 80, 376–383. https://doi.org/10.2527/ Automatic early warning of tail biting in pigs: 3D cameras can detect lowered tail 2002.802376x. posture before an outbreak. PLoS One 13, 18. https://doi.org/10.1371/journal. Gertheiss, J., Maier, V., Hessel, E.F., Staicu, A.M., 2015. Marginal functional regression pone.0194524. models for analyzing the feeding behavior of pigs. J. Agric. Biol. Environ. Stat. 20, de Haer, L.C.M., de Vries, A.G., 1993. Feed intake patterns of and feed digestibility in 353–370. https://doi.org/10.1007/s13253-015-0212-7. growing pigs housed individually or in groups. Livest. Prod. Sci. 33, 277–292. Godcharles, L., Robert, S., Matte, J.J., Bertin-Mahieux, J., Martineau, G., 1993. Transient https://doi.org/10.1016/0301-6226(93)90008-6. stray voltage: is it detrimental to growth performance, health status and welfare of de Haer, L.C.M., Merks, J.W.M., Kooper, H.G., Buiting, G.A.J., van Hattum, J.Â., 1992. market pigs? Vet. Res. Commun. 17, 41–53. A note on the IVOG®-station: a feeding station to record the individual feed intake of Gonyou, H.W., Lou, Z., 2000. Effects of eating space and availability of water in feeders group housed growing pigs. Relevance of Eating Pattern for Selection of Growing on productivity and eating behavior of grower/finisher pigs. J. Anim. Sci. 78, Pigs. Landbouwuniversiteit te Wageningen, Wageningen, pp. 11–20 (PhD Thesis). 865–870. https://doi.org/10.2527/2000.784865x. De Haer, L., Merks, J., 1992. Patterns of daily feed intake in growing pigs. Relevance of Gonyou, H.W., Chapple, R.P., Frank, G.R., 1992. Productivity, time budgets and social Eating Pattern for Selection of Growing Pigs. Landbouwuniversiteit te Wageningen, aspects of eating in pigs penned in groups of five or individually. Appl. Anim. Behav. Wageningen, Netherlands, pp. 19–44 (PhD Thesis). Sci. 34, 291–301. https://doi.org/10.1016/S0168-1591(05)80090-5. de Vries, M., Bokkers, E.A.M., van Schaik, G., Botreau, R., Engel, B., Dijkstra, T., de Green, T.C., Mellor, D.J., 2011. Extending ideas about animal welfare assessment to Boer, I.J.M., 2013. Evaluating results of the Welfare Quality multi-criteria evaluation include “quality of life” and related concepts. N. Z. Vet. J. 59, 263–271. https://doi. model for classification of dairy cattle welfare at the herd level. J. Dairy Sci. 96, org/10.1080/00480169.2011.610283. 6264–6273. Guarino, M., Jans, P., Costa, A., Aerts, J.M., Berckmans, D., 2008. Field test of algorithm Ding, R., Quan, J., Yang, M., Wang, X., Zheng, E., Yang, H., Fu, D., Yang, Y., Yang, L., for automatic cough detection in pig houses. Comput. Electron. Agric. 62, 22–28. Li, Z., Liu, D., Cai, G., Wu, Z., Yang, J., 2017. Genome-wide association analysis https://doi.org/10.1016/j.compag.2007.08.016. reveals genetic loci and candidate genes for feeding behavior and eating efficiency in Guo, Y., Lian, X., Yan, P., 2015a. Diurnal rhythms, locations and behavioural sequences Duroc boars. PLoS One 12, 1–15. https://doi.org/10.1371/journal.pone.0183244. associated with eliminative behaviours in fattening pigs. Appl. Anim. Behav. Sci. Ding, R., Yang, M., Wang, X., Quan, J., Zhuang, Z., Zhou, S., Li, S., Xu, Z., Zheng, E., 168, 18–23. https://doi.org/10.1016/j.applanim.2015.01.011. Cai, G., Liu, D., Huang, W., Yang, J., Wu, Z., 2018. Genetic architecture of feeding Guo, Y.M., Zhang, Z.Y., Ma, J.W., Ai, H.S., Ren, J., Huang, L.S., 2015b. A genomewide behavior and feed efficiency in a Duroc pig population. Front. Genet. 9, 11. https:// association study of feed efficiency and feeding behaviors at two fattening stages in a doi.org/10.3389/fgene.2018.00220. White Duroc × Erhualian F2 population. J. Anim. Sci. 93, 1481–1489. https://doi. Do, D.N., Strathe, A.B., Jensen, J., Mark, T., Kadarmideen, H.N., 2013a. Genetic org/10.2527/jas.2014-8655. parameters for different measures of feed efficiency and related traits in boars of Hall, A.D., Hill, W.G., Bampton, P.R., Webb, A.J., 1999a. Predicted responses to selection three pig breeds. J. Anim. Sci. 91, 4069–4079. https://doi.org/10.2527/jas.2012- from indices incorporating feeding pattern traits of pigs using electronic feeders. 6197. Anim. Sci. 68, 407–412. https://doi.org/10.1017/S1357729800050402. Do, D.N., Strathe, A.B., Ostersen, T., Jensen, J., Mark, T., Kadarmideen, H.N., 2013b. Hall, A.D., Hill, W.G., Bampton, P.R., Webb, A.J., 1999b. Genetic and phenotypic Genome-wide association study reveals genetic architecture of eating behavior in parameter estimates for feeding pattern and performance test traits in pigs. Anim. pigs and its implications for humans obesity by comparative mapping. PLoS One 8, Sci. 68, 43–48. https://doi.org/10.1017/s1357729800050062. 14. https://doi.org/10.1371/journal.pone.0071509. Hansen, L.L., Hagelsø, A.M., Madsen, A., 1982. Behavioural results and performance of dos Santos, L.S., Pomar, C., Campos, P.H.R.F., da Silva, W.C., Gobi, J., de, P., Veira, A.M., bacon pigs fed “ad libitum” from one or several self-feeders. Appl. Anim. Ethol. 8, Fraga, A.Z., Hauschild, L., 2018. Precision feeding strategy for growing pigs under 307–333. https://doi.org/10.1016/0304-3762(82)90065-7. heat stress conditions. J. Anim. Sci. 96, 4789–4801. https://doi.org/10.1093/jas/ Hauschild, L., Kristensen, A.R., Andretta, I., Remus, A., Santos, L.S., Pomar, C., 2020. sky343. Toward better estimates of the real-time individual amino acid requirements of Edmonds, M.S., Arentson, B.E., Mente, G.A., 1998. Effect of protein levels and space growing-finishing pigs showing deviations from their typical feeding patterns. allocations on performance of growing-finishing pigs. J. Anim. Sci. 76, 814–821. Animal 14 (2), s371–s381. https://doi.org/10.1017/S1751731120001226. https://doi.org/10.2527/1998.763814x. Heckmann, S., Mannl, R., Hoy, S., 2018. Investigations on possible relationships between Edwards, S.A., Armsby, A.W., Spechter, H.H., 1988. Effects of floor area allowance on caudophagia and the rank position of fattening pigs measured by an electronic performance of growing pigs kept on fully slatted floors. Anim. Prod. 46, 453–459. feeding station. Zuchtungskunde 90, 353–363. https://doi.org/10.1017/S0003356100019061. Helm, E.T., Outhouse, A.C., Schwartz, K.J., Dekkers, J.C.M., Lonergan, S.M., Rauw, W. Eigenberg, R.A., Nienaber, J.A., Hahn, G.L.R., Kachman, S.D., 2002. Swine response to M., Gabler, N.K., 2018a. Impact of Mycoplasma hyopneumoniae and lawsonia misting synchronized with meal events. Appl. Eng. Agric. 18, 347–350. intracellularis on the performance of pigs divergently selected for feed efficiency. Er, C., Lium, B., Tavornpanich, S., Hofmo, P.O., Forberg, H., Hauge, A.G., Grøntvedt, C. J. Anim. Sci. 96, 462–472. https://doi.org/10.1093/jas/skx074. A., Framstad, T., Brun, E., 2014. Adverse effects of Influenza A(H1N1)pdm09 virus Helm, E.T., Outhouse, A.C., Schwartz, K.J., Lonergan, S.M., Curry, S.M., Dekkers, J.C.M., infection on growth performance of Norwegian pigs - a longitudinal study at a boar Gabler, N.K., 2018b. Metabolic adaptation of pigs to a Mycoplasma hyopneumoniae testing station. BMC Vet. Res. 10, 1–14. https://doi.org/10.1186/s12917-014-0284- and Lawsonia intracellularis dual challenge. J. Anim. Sci. 96, 3196–3207. https:// 6. doi.org/10.1093/jas/sky220. Fàbrega, E., Tibau, J., Soler, J., Fernández, J., Font, J., Carrión, D., Diestre, A., Herrera-Cáceres, W., Ragab, M., Sánchez, J.P., 2019. Indirect genetic effects on the Manteca, X., 2003. Feeding patterns, growth performance and carcass traits in relationships between production and feeding behaviour traits in growing Duroc group-housed growing-finishing pigs: the effect of terminal sire line, halothane pigs. Animal 14, 233–242. https://doi.org/10.1017/S1751731119002179. genotype and age. Anim. Sci. 77, 11–21. https://doi.org/10.1017/ Howie, J.A., Tolkamp, B.J., Avendaño, S., Kyriazakis, I., 2009. A novel flexible method to s1357729800053601. split feeding behaviour into bouts. Appl. Anim. Behav. Sci. 116, 101–109. https:// Fàbrega, E., Velarde, A., Cros, J., Gispert, M., Suárez, P., Tibau, J., Soler, J., 2010. Effect doi.org/10.1016/j.applanim.2008.09.005. of vaccination against gonadotrophin-releasing hormone, using Improvac®, on Hoy, S., Schamun, S., Weirich, C., 2012. Investigations on feed intake and social growth performance, body composition, behaviour and acute phase proteins. Livest. behaviour of fattening pigs fed at an electronic feeding station. Appl. Anim. Behav. Sci. 132, 53–59. https://doi.org/10.1016/j.livsci.2010.04.021. Sci. 139, 58–64. https://doi.org/10.1016/j.applanim.2012.03.010. Feddes, J.J.R., Young, B.A., DeShazer, J.A., 1989. Influence of temperature and light on Hsia, L.C., Wood-Gush, D.G.M., 1983. A note on social facilitation and competition in the feeding behaviour of pigs. Appl. Anim. Behav. Sci. 23, 215–222. https://doi.org/ feeding behaviour of pigs. Anim. Prod. 37, 149–152. https://doi.org/10.1017/ 10.1016/0168-1591(89)90112-3. S0003356100001677. Fernández, J., Fàbrega, E., Soler, J., Tibau, J., Ruiz, J.L., Puigvert, X., Manteca, X., 2011. Hsia, L.C., Wood-Gush, D.G.M., 1984a. The temporal patterns of food intake and Feeding strategy in group-housed growing pigs of four different breeds. Appl. Anim. allelomimetic feeding by pigs of different ages. Appl. Anim. Ethol. 11, 271–282. Behav. Sci. 134, 109–120. https://doi.org/10.1016/j.applanim.2011.06.018. https://doi.org/10.1016/0304-3762(84)90034-8. Fraga, A.Z., Campos, P.H.R.F., da Silva, W.C., Caetano, R.P., Veira, A.M., dos Santos, L.S., Hsia, L.C., Wood-Gush, D.G.M., 1984b. Social facilitation in the feeding behaviour of pigs Hauschild, L., 2019. Sequential feeding with high-fat/low-crude protein diets for two and the effect of rank. Appl. Anim. Ethol. 11, 265–270. 14 J.D. Bus et al. Applied Animal Behaviour Science 241 (2021) 105383 Hyun, Y., Ellis, M., 2000. Relationships between feed intake traits, monitored using a pigs. Comput. Electron. Agric. 102, 10–18. https://doi.org/10.1016/j. computerized feed intake recording system, and growth performance and body compag.2013.12.015. composition of group-housed pigs. Asian-Australas. J. Anim. Sci. 13, 1717–1725. Maselyne, J., Saeys, W., Van Nuffel, A., 2015. Review: quantifying animal feeding Hyun, Y., Ellis, M., 2001. Effect of group size and feeder type on growth performance and behaviour with a focus on pigs. Physiol. Behav. 138, 37–51. https://doi.org/ feeding patterns in growing pigs. J. Anim. Sci. 79, 803–810. https://doi.org/ 10.1016/j.physbeh.2014.09.012. 10.2527/2001.794803x. Maselyne, J., Van Nuffel, A., Briene, P., Vangeyte, J., De Ketelaere, B., Millet, S., Van den Hyun, Y., Ellis, M., 2002. Effect of group size and feeder type on growth performance and Hof, J., Maes, D., Saeys, W., 2018. Online warning systems for individual fattening feeding patterns in finishing pigs. J. Anim. Sci. 80, 568–574. https://doi.org/ pigs based on their feeding pattern. Biosyst. Eng. 173, 143–156. https://doi.org/ 10.2527/2002.803568x. 10.1016/j.biosystemseng.2017.08.006. Hyun, Y., Ellis, M., McKeith, F.K., Wilson, E.R., 1997. Feed intake pattern of group- McGlone, J.J., Newby, B.E., 1994. Space requirements for finishing pigs in confinement: housed growing-finishing pigs monitored using a computerized feed intake behavior and performance while group size and space vary. Appl. Anim. Behav. Sci. recording system. J. Anim. Sci. 75, 1443–1451. https://doi.org/10.2527/ 39, 331–338. https://doi.org/10.1016/0168-1591(94)90166-X. 1997.7561443x. Mensching, A., Henne, H., Simianer, H., Sharifi, A., 2018. Analysis of food intake Hyun, Y., Ellis, M., Johnson, R.W., 1998a. Effects of feeder type, space allowance, and associated behavioral traits of boars using data from transponder feeding. mixing on the growth performance and feed intake pattern of growing pigs. J. Anim. Zuchtungskunde 90, 57–71. Sci. 76, 2771–2778. Mikesell, R.E., Kephart, K.B., 1999. Effect of grouping arrangement on behavior and Hyun, Y., Ellis, M., Riskowski, G., Johnson, R., 1998b. Growth performance of pigs performance of finishing pigs. Livest. Prod. Sci. 57, 291–294. https://doi.org/ subjected to multiple concurrent environmental stressors. J. Anim. Sci. 76, 721–727. 10.1016/S0301-6226(98)00179-1. https://doi.org/10.2527/1998.763721x. Montgomery, G.W., Flux, D.S., Carr, J.R., 1978. Feeding patterns in pigs: the effects of Hyun, Y., Wolter, B., Ellis, M., 2001. Feed intake patterns and growth performance of amino acid deficiency. Physiol. Behav. 20, 693–698. https://doi.org/10.1016/0031- purebred and crossbred Meishan and Yorkshire pigs. Anim. Sci. 14, 837–843. 9384(78)90294-9. Ingram, D.L., Walters, D.E., Legge, K.F., 1980. Variations in motor activity and in food Morales, A., Ibarra, N., Chávez, M., Gómez, T., Suárez, A., Valle, J.A., Camacho, R.L., and water intake over 24h periods in pigs. J. Agric. Sci. 95, 371–380. https://doi. Cervantes, M., 2018. Effect of feed intake level and dietary protein content on the org/10.1017/S002185960003940X. body temperature of pigs housed under thermo neutral conditions. J. Anim. Physiol. Jensen, M.B., Pedersen, L.J., 2010. Effects of feeding level and access to rooting material Anim. Nutr. (Berl.) 102, e718–e725. https://doi.org/10.1111/jpn.12824. on behaviour of growing pigs in situations with reduced feeding space and delayed Morgan, C.A., Deans, L.A., Lawrence, A.B., Nielsen, B.L., 1998. The effects of straw feeding. Appl. Anim. Behav. Sci. 123, 1–6. https://doi.org/10.1016/j. bedding on the feeding and social behaviour of growing pigs fed by means of single- applanim.2009.12.015. space feeders. Appl. Anim. Behav. Sci. 58, 23–33. https://doi.org/10.1016/S0168- Jensen, M.B., Herskin, M.S., Canibe, N., Forkman, B., Pedersen, L.J., 2019. Effect of straw 1591(97)00141-X. amount on feed intake and weight gain in growing pigs housed in pens with partly Morgan, C.A., Deans, L.A., Lawrence, A.B., Nielsen, B.L., 1999. The effects of changing slatted floor. Animal 1–8. https://doi.org/10.1017/S1751731119002957. straw provision on the feeding behaviour and activity of growing pigs given food Jiao, S., 2015. Genome-wide association for different measures of feed efficiency and through single-space feeders. Anim. Sci. 68, 19–26. https://doi.org/10.1017/ feeding behavior. The Genetics of Feed Efficiency and Feeding Behavior in Pigs s1357729800050037. 209–222. Morgan, C.A., Emmans, G.C., Tolkamp, B.J., Kyriazakis, I., 2000. Analysis of the feeding Johnson, R.W., 2002. The concept of sickness behavior: a brief chronological account of behavior of pigs using different models. Physiol. Behav. 68, 395–403. https://doi. four key discoveries. Vet. Immunol. Immunopathol. 87, 443–450. https://doi.org/ org/10.1016/S0031-9384(99)00195-X. 10.1016/S0165-2427(02)00069-7. Morrison, R.S., Hemsworth, P.H., Cronin, G.M., Campbell, R.G., 2003. The effect of Kapun, A., Adrion, F., Jungbluth, T., 2016. Test of a UHF-RFID system for health restricting pen space and feeder availability on the behaviour and growth monitoring of finishing pigs. International Conference of Agricultural Engineering. performance of entire male growing pigs in a deep-litter, large group housing Kapun, A., Adrion, F., Gallmann, E., 2017. Behaviour and activity monitoring of growing- system. Appl. Anim. Behav. Sci. 83, 163–176. https://doi.org/10.1016/S0168-1591 finishing pigs with UHF-RFID. In: 8th European Conference on Precision Livestock (03)00117-5. Farming. ECPLF, pp. 605–613. Morrison, R.S., Johnston, L.J., Hilbrands, A.M., 2007. The behaviour, welfare, growth Kavlak, A.T., Uimari, P., 2019. Estimation of heritability of feeding behaviour traits and performance and meat quality of pigs housed in a deep-litter, large group housing their correlation with production traits in Finnish Yorkshire pigs. J. Anim. Breed. system compared to a conventional confinement system. Appl. Anim. Behav. Sci. Genet. 136, 484–494. https://doi.org/10.1111/jbg.12408. 103, 12–24. https://doi.org/10.1016/j.applanim.2006.04.002. Kerr, C.A., Giles, L.R., Jones, M.R., Reverter, A., 2005. Effects of grouping unfamiliar Morrow, A.T.S., Walker, N., 1994a. Effects of number and siting of single-space feeders cohorts, high ambient temperature and stocking density on live performance of on performance and feeding behaviour of growing pigs. J. Agric. Sci. 122, 465–470. growing pigs. J. Anim. Sci. 83, 908–915. https://doi.org/10.2527/2005.834908x. https://doi.org/10.1017/S002185960006740X. Labroue, F., Guéblez, R., Sellier, P., Meunier-Salaün, M.C., 1994. Feeding behaviour of Morrow, A.T.S., Walker, N., 1994b. A note on changes to feeding behaviour of growing group-housed Large White and Landrace pigs in French central test stations. Livest. pigs by fitting stalls to single-space feeders. Anim. Prod. 59, 151–153. https://doi. Prod. Sci. 40, 303–312. https://doi.org/10.1016/0301-6226(94)90097-3. org/10.1017/S0003356100007613. Labroue, F., Guéblez, R., Sellier, P., 1997. Genetic parameters of feeding behaviour and Munsterhjelm, C., Heinonen, M., Valros, A., 2015. Effects of clinical lameness and tail performance traits in group-housed Large White and French Landrace growing pigs. biting lesions on voluntary feed intake in growing pigs. Livest. Sci. 181, 210–219. Genet. Sel. Evol. 29, 451–468. https://doi.org/10.1051/gse:19970403. https://doi.org/10.1016/j.livsci.2015.09.003. Labroue, F., Guéblez, R., Meunier-Salaün, M.C., Sellier, P., 1999. Feed intake behaviour Munsterhjelm, C., Nordgreen, J., Aae, F., Heinonen, M., Olstad, K., Aasmundstad, T., of group-housed Piétrain and Large White growing pigs. Ann. Zootech. 48, 247–261. Janczak, A.M., Valros, A., 2017. To be blamed or pitied? The effect of illness on https://doi.org/10.1051/animres:19990402. social behavior, cytokine levels and feed intake in undocked boars. Physiol. Behav. Laitat, M., Antoine, N., Cabaraux, J.F., Cassart, D., Mainil, J., Moula, N., Nicks, B., 179, 298–307. https://doi.org/10.1016/j.physbeh.2017.06.024. Wavreille, J., Philippe, F.X., 2015. Influence of sugar beet pulp on feeding behavior, National Research Council, 2012. Nutrient Requirements of Swine, 11th ed. The National growth performance, carcass quality and gut health of fattening pigs. Biotechnol. Academies Press, Washington, DC. https://doi.org/10.17226/13298. Agron. Soc. Environ. 19, 20–31. Nielsen, B.L., 1999. On the interpretation of feeding behaviour measures and the use of Leiber-Schotte, C., 2009. Einfluss der Rangordnung bei Jungebern im Eigenleistungstest feeding rate as an indicator of social constraint. Appl. Anim. Behav. Sci. 63, 79–91. auf Futteraufnahme und Futteraufnahmeverhalten unter Berücksichtigung https://doi.org/10.1016/S0168-1591(99)00003-9. endokrinologischer und immunologischer Parameter. Christian-Albrechts- Nielsen, B.L., Lawrence, A.B., Whittemore, C.T., 1995. Effects of group size on feeding Universität zu Kiel. behaviour, social behaviour, and performance of growing pigs using single-space Li, Y.Z., Wang, L.H., Johnston, L.J., 2012. Effects of farrowing system on behavior and feeders. Livest. Prod. Sci. 44, 73–85. growth performance of growing-finishing pigs. J. Anim. Sci. 90, 1008–1014. https:// Nielsen, B.L., Lawrence, A.B., Whittemore, C.T., 1996a. Feeding behaviour of growing doi.org/10.2527/jas.2011-4050. pigs using single or multi-space feeders. Appl. Anim. Behav. Sci. 47, 235–246. Lopez, J., Jesse, G., Becker, B., Ellersieck, M., 1991a. Effects of temperature on the https://doi.org/10.1016/0168-1591(95)00649-4. performance of finishing swine: I. Effects of a hot, diurnal temperature on average Nielsen, B.L., Lawrence, A.B., Whittemore, C.T., 1996b. Effect of individual housing on daily gain, feed intake, and feed efficiency. J. Anim. Sci. 69, 1843–1849. https://doi. the feeding behaviour of previously group housed growing pigs. Appl. Anim. Behav. org/10.2527/1991.6951843x. Sci. 47, 149–161. https://doi.org/10.1016/0168-1591(95)01001-7. Lopez, J., Jesse, G., Becker, B.A., Ellersieck, M.R., 1991b. Effects of temperature on the Nienaber, J.A., Hahn, G.L.R., Yen, J.T., 1987. Thermal environment effects on growing- performance of finishing swine: II. Effects of a cold, diurnal temperature on average finishing swine: part I - growth, feed intake and heat production. Trans. Am. Soc. daily gain, feed intake, and feed efficiency. J. Anim. Sci. 69, 1850–1855. https://doi. Agric. Eng. 30, 1772–1775. https://doi.org/10.13031/2013.30635. org/10.2527/1991.6951850x. Nienaber, J.A., McDonald, T.P., Hahn, G.L., Chen, Y.R., 1990. Eating dynamics of Lorenzo Bermejo, J., Roehe, R., Rave, G., Kalm, E., 2003. Comparison of linear and growing-finishing swine. Transactions of the American Society of Agricultural nonlinear functions and covariance structures to estimate feed intake pattern in Engineers, pp. 2011–2018. https://doi.org/10.13031/2013.31571. growing pigs. Livest. Prod. Sci. 82, 15–26. https://doi.org/10.1016/S0301-6226(03) Nienaber, J., McDonald, T., Hahn, G., Chen, Y., 1991. Group feeding behavior of swine. 00005-8. Trans. Am. Soc. Agric. Eng. 34, 289–294 (Abstr.). Lu, D., Jiao, S., Tiezzi, F., Knauer, M., Huang, Y., Gray, K.A., Maltecca, C., 2017. The Nienaber, J.A., Hahn, G.L., Mcdonald, T.P., Korthals, R.L., 1996. Feeding patterns and relationship between different measures of feed efficiency and feeding behavior swine performance in hot environments. Trans. Am. Soc. Agric. Eng. 39, 195–202. traits in Duroc pigs. J. Anim. Sci. 95, 3370. https://doi.org/10.2527/jas2017.1509. O’Connell, N.E., Beattie, V.E., Weatherup, R.N., 2004. Influence of group size during the Maselyne, J., Saeys, W., De Ketelaere, B., Mertens, K., Vangeyte, J., Hessel, E.F., post-weaning period on the performance and behaviour of pigs. Livest. Prod. Sci. 86, Millet, S., Van Nuffel, A., 2014. Validation of a high frequency radio frequency 225–232. https://doi.org/10.1016/j.livprodsci.2003.07.001. identification (HF RFID) system for registering feeding patterns of growing-finishing 15 J.D. Bus et al. Applied Animal Behaviour Science 241 (2021) 105383 Parois, S., Larzul, C., Prunier, A., 2017. Associations between the dominance status and Schouten, W.G.P., 1986. Rearing Conditions and Behaviour in Pigs. sexual development, skin lesions or feeding behaviour of intact male pigs. Appl. Landbouwhogeschool Wageningen. Anim. Behav. Sci. 187, 15–22. https://doi.org/10.1016/j.applanim.2016.12.001. Schulze, V., Roehe, R., Looft, H., Kalm, E., 2001. Effects of continuous and periodic Pastorelli, H., Van Milgen, J., Lovatto, P., Montagne, L., 2012. Meta-analysis of feed feeding by electronic feeders on accuracy of measuring feed intake information and intake and growth responses of growing pigs after a sanitary challenge. Animal 6, their genetic association with growth performances. J. Anim. Breed. Genet. 118, 952–961. https://doi.org/10.1017/S175173111100228X. 403–416. https://doi.org/10.1046/j.1439-0388.2001.00158.x. Pearce, G.P., Paterson, A.M., 1993. The effect of space restriction and provision of toys Schweer, W., Schwartz, K., Burrough, E., Yoon, K., Sparks, J., Gabler, N., 2016. The effect during rearing on the behaviour, productivity and physiology of male pigs. Appl. of porcine reproductive and respiratory syndrome virus and porcine epidemic Anim. Behav. Sci. 36, 11–28. https://doi.org/10.1016/0168-1591(93)90095-7. diarrhea virus challenge on growing pigs I: growth performance and digestibility. Peeters, E., Driessen, B., Moons, C.P.H., Ödberg, F.O., Geers, R., 2006. Effect of J. Anim. Sci. 94, 514–522. https://doi.org/10.2527/jas2015-9834. temporary straw bedding on pigs’ behaviour, performance, cortisol and meat Sell-Kubiak, E., van der Waaij, E.H., Bijma, P., 2013. Effect of gestating sow body quality. Appl. Anim. Behav. Sci. 98, 234–248. https://doi.org/10.1016/j. condition, feed refusals, and group housing on growth and feed intake in grower- applanim.2005.10.002. finishing pigs. J. Anim. Sci. 91, 3538–3548. https://doi.org/10.2527/jas.2012-6027. Puppe, B., Hoy, S., Jakob, M., Wullbrandt, H., 1991. Preliminary results obtained from Simonsen, H.B., 1990. Behaviour and distribution of fattening pigs in the multi-activity studies into social order of female and male fattening pigs with sexes kept together pen. Appl. Anim. Behav. Sci. 27, 311–324. https://doi.org/10.1016/0168-1591(90) and separate - correlations with live weight development and selected behaviors. 90127-Y. Monatsh. Veterinarmed. 46, 515–519 (Abstr.). Spoolder, H.A.M., Edwards, S.A., Corning, S., 1999. Effects of group size and feeder space Putz, A.M., Harding, J.C.S., Dyck, M.K., Fortin, F., Plastow, G.S., Dekkers, J.C.M., PigGen allowance on welfare in finishing pigs. Anim. Sci. 69, 481–489. https://doi.org/ Canada, 2019. Novel resilience phenotypes using feed intake data from a natural 10.1017/S135772980005133X. disease challenge model in wean-to-finish pigs. Front. Genet. 10, 1–14. https://doi. Tallet, C., Brilloüet, A., Meunier-Salaün, M.C., Paulmier, V., Guérin, C., Prunier, A., 2013. org/10.3389/fgene.2018.00660. Effects of neonatal castration on social behaviour, human-animal relationship and Quemeneur, K., Labussiere, E., Le Gall, M., Lechevestrier, Y., Montagne, L., 2019. feeding activity in finishing pigs reared in a conventional or an enriched housing. Feeding behaviour and pre-prandial status affect post-prandial plasma energy Appl. Anim. Behav. Sci. 145, 70–83. https://doi.org/10.1016/j. metabolites and insulin kinetics in growing pigs fed diets differing in fibre applanim.2013.03.001. concentration. Br. J. Nutr. 121, 625–636. https://doi.org/10.1017/ Thomsen, L.R., Nielsen, B.L., Larsen, O.N., 2009. Implications of food patch distribution S0007114518003768. on social foraging in domestic pigs (Sus scrofa). Appl. Anim. Behav. Sci. 122, Quiniou, N., Noblet, J., 2012. Effect of the dietary net energy concentration on feed 111–118. https://doi.org/10.1016/j.applanim.2009.11.006. intake and performance of growing-finishing pigs housed individually. J. Anim. Sci. Tindsley, W.E.C., Lean, I.J., 1984. Effects of weight range at allocation on production and 90, 4362–4372. https://doi.org/10.2527/jas2011-4004. behaviour in fattening pig groups. Appl. Anim. Behav. Sci. 12, 79–92. https://doi. Quiniou, N., Dubois, S., Noblet, J., 2000a. Voluntary feed intake and feeding behaviour org/10.1016/0168-1591(84)90098-4. of group-housed growing pigs are affected by ambient temperature and body weight. Toft, N., Madsen, T., 1998. Modeling eating patterns of growing pigs using dynamic Livest. Prod. Sci. 63, 245–253. https://doi.org/10.1016/S0301-6226(99)00135-9. linear models. Dina Research Report No. 69. Quiniou, N., Massabie, P., Granier, R., 2000b. Diurnally variation of ambient Turner, S.P., Dahlgren, M., Arey, D.S., Edwards, S.A., 2002. Effect of social group size temperature around 24 or 28◦ C: influence on performance and feeding behavior of and initial live weight on feeder space requirement of growing pigs given food ad growing pigs. Swine Housing Conference. American Society of Agricultural and libitum. Anim. Sci. 75, 75–83. https://doi.org/10.1017/S1357729800052851. Biological Engineers. https://doi.org/10.13031/2013.110, p. 8. Viitasaari, E., Raekallio, M., Valros, A., Peltoniemi, O., Hänninen, L., Heinonen, M., Ragab, M., Piles, M., Quintanilla, R., Sánchez, J.P., 2019. Indirect genetic effect model 2015. The effect of ketoprofen on feeding behavior of tail-bitten pigs. Porc. Heal. using feeding behaviour traits to define the degree of interaction between mates: An Manag. 1, 1–7. https://doi.org/10.1186/s40813-015-0005-y. implementation in pigs growth rate. Animal 13, 231–239. https://doi.org/10.1017/ Villagrá, A., Althaus, R.L., Lainez, M., Martinez, A.B., Torres, A.G., 2007. Modelling of S1751731118001192. daily rhythms of behavioural patterns in growing pigs on two commercial farms. Rasmussen, D.K., Weber, R., Wechsler, B., 2006a. Performance, lean meat proportion and Biol. Rhythm Res. 38, 347–354. https://doi.org/10.1080/09291010600950131. behaviour of fattening pigs given a liquid diet at different animal/feeding-place Von Felde, A., Roehe, R., Looft, H., Kalm, E., 1996. Genetic association between feed ratios. Anim. Sci. 82, 575–580. https://doi.org/10.1079/ASC200668. intake and feed intake behaviour at different stages of growth of group-housed boars. Rasmussen, D.K., Weber, R., Wechsler, B., 2006b. Effects of animal/feeding-place ratio Livest. Prod. Sci. 47, 11–22. https://doi.org/10.1016/S0301-6226(96)01006-8. on the behaviour and performance of fattening pigs fed via sensor-controlled liquid Walker, N., 1991. The effects on performance and behaviour of number of growing pigs feeding. Appl. Anim. Behav. Sci. 98, 45–53. https://doi.org/10.1016/j. per mono-place feeder. Anim. Feed Sci. Technol. 35, 3–13. https://doi.org/10.1016/ applanim.2005.08.008. 0377-8401(91)90094-9. Rauw, W.M., Soler, J., Tibau, J., Reixach, J., Raya, L.G., 2006. The relationship between Wallenbeck, A., Keeling, L.J., 2013. Using data from electronic feeders on visit frequency residual feed intake and feed intake behavior in group-housed Duroc barrows. and feed consumption to indicate tail biting outbreaks in commercial pig production. J. Anim. Sci. 84, 956–962. https://doi.org/10.2527/2006.844956x. J. Anim. Sci. 91, 2879–2884. https://doi.org/10.2527/jas.2012-5848. Renaudeau, D., 2009. Effect of housing conditions (clean vs. dirty) on growth Webb, L.E., Veenhoven, R., Harfeld, J.L., Jensen, M.B., 2019. What is animal happiness? performance and feeding behavior in growing pigs in a tropical climate. Trop. Anim. Ann. N. Y. Acad. Sci. 1438, 62–76. https://doi.org/10.1111/nyas.13983. Health Prod. 41, 559–563. https://doi.org/10.1007/s11250-008-9223-5. Wei, S., Guo, Y., Yan, P., 2019. Comparison of two housing systems on behaviour and Renaudeau, D., Siloux, F., Giorgi, M., Weisbecker, J., 2005. A comparison of growth performance of fattening pigs. J. Appl. Anim. Res. 47, 41–45. https://doi.org/ performance and feeding behaviour in Creole and Large White pigs: preliminary 10.1080/09712119.2018.1561372. results. Arch. Zootec. 54, 471–476. Weiler, U., Götz, M., Schmidt, A., Otto, M., Müller, S., 2013. Influence of sex and Renaudeau, D., Giorgi, M., Silou, F., Weisbecker, J.L., 2006. Effect of breed (lean or fat immunocastration on feed intake behavior, skatole and indole concentrations in pigs) and sex on performance and feeding behaviour of group housed growing pigs in adipose tissue of pigs. Animal 7, 300–308. https://doi.org/10.1017/ a tropical climate. Asian-Australas. J. Anim. Sci. 19, 593–600. https://doi.org/ S175173111200167X. 10.5713/ajas.2006.593. Welfare Quality®, 2009. Welfare Quality® Assessment Protocol for Pigs (Sows and Reyer, H., Shirali, M., Ponsuksili, S., Murani, E., Varley, P.F., Jensen, J., Wimmers, K., Piglets, Growing and Finishing Pigs). Welfare Quality® Consortium, Lelystad, 2017. Exploring the genetics of feed efficiency and feeding behaviour traits in a pig Netherlands. Lelystad, the Netherlands. line highly selected for performance characteristics. Mol. Genet. Genomics 292, Wellock, I.J., Emmans, G.C., Kyriazakis, I., 2003. Predicting the consequences of social 1001–1011. https://doi.org/10.1007/s00438-017-1325-1. stressors on pig food intake and performance. J. Anim. Sci. 81, 2995–3007. https:// Rinaldo, D., Le Dividich, J., Noblet, J., 2000. Adverse effects of tropical climate on doi.org/10.2527/2003.81122995x. voluntary feed intake and performance of growing pigs. Livest. Prod. Sci. 66, Whittemore, C.T., Green, D.M., Knap, P.W., 2001a. Technical review of the energy and 223–234. https://doi.org/10.1016/S0301-6226(00)00181-0. protein requirements of growing pigs: food intake. Anim. Sci. 73, 3–17. https://doi. Rohrer, G.A., Brown-Brandl, T., Rempel, L.A., Schneider, J.F., Holl, J., 2013. Genetic org/10.1017/S1357729800058008. analysis of behavior traits in swine production. Livest. Sci. 157, 28–37. https://doi. Whittemore, E.C., Kyriazakis, I., Emmans, G.C., Tolkamp, B.J., 2001b. Tests of two org/10.1016/j.livsci.2013.07.002. theories of food intake using growing pigs 1. The effect of ambient temperature on Sandberg, F.B., Emmans, G.C., Kyriazakis, I., 2006. A model for predicting feed intake of the intake of foods of differing bulk content. Anim. Sci. 72, 351–360. https://doi. growing animals during exposure to pathogens. J. Anim. Sci. 84, 1552–1566. org/10.1017/s1357729800055855. https://doi.org/10.2527/2006.8461552x. Wiegand, R.M., Gonyou, H.W., Curtis, S.E., 1994. Pen shape and size: effects on pig Schamun, S., Hoy, S., 2011. Untersuchungen zum Futteraufnahmeverhalten von behavior and performance. Appl. Anim. Behav. Sci. 39, 49–61. https://doi.org/ Mastschweinen an einer elektronischen Abrufstation unter Berücksichtigung der 10.1016/0168-1591(94)90015-9. Rangordnung. Züchtungskunde 83, S426–S438. Xin, H., DeShazer, J.A., 1992. Feeding patterns of growing pigs at warm constant and Schmidt, T., Calabrese, J.M., Grodzycki, M., Paulick, M., Pearce, M.C., Rau, F., von cyclic temperatures. Trans. Am. Soc. Agric. Eng. 35, 319–323. https://doi.org/ Borell, E., 2011. Impact of single-sex and mixed-sex group housing of boars 10.13031/2013.28606. vaccinated against GnRF or physically castrated on body lesions, feeding behaviour Xin, W., Li, X., Zhang, F., Yan, G., Ding, N., Huang, L., Zhang, Z., 2016. A multi- and weight gain. Appl. Anim. Behav. Sci. 130, 42–52. https://doi.org/10.1016/j. population survey on swine feeding behavior with electronic feeding devices. Arch. applanim.2010.11.019. Anim. Breed. 59, 445–452. https://doi.org/10.5194/aab-59-445-2016. Schmolke, S.A., Li, Y.Z., Gonyou, H.W., 2003. Effect of group size on performance of Yeates, J.W., 2011. Is “a life worth living” a concept worth having? Anim. Welf. 20, growing-finishing pigs. J. Anim. Sci. 81, 874–878. https://doi.org/10.2527/ 397–406. https://doi.org/10.1111/jpc.12550. 2003.814874x. Yeates, J.W., Main, D.C.J., 2008. Assessment of positive welfare: a review. Vet. J. 175, Schönfelder, A., 2005. The effect of rank order on feed intake and growth of fattening 293–300. https://doi.org/10.1016/j.tvjl.2007.05.009. pigs. Dtsch. Tierarztl. Wochenschr. 112, 215–218 (Abstr.). 16 J.D. Bus et al. Applied Animal Behaviour Science 241 (2021) 105383 Young, R.J., Lawrence, A.B., 1994. Feeding behaviour of pigs in groups monitored by a Zeng, Z., Huan, Y., Zhu, L., Chen, Z., Yuan, Z., Zeng, H., Sun, D., Wang, Z., Yan, Z., 2019. computerized feeding system. Anim. Prod. 58, 145–152. https://doi.org/10.1017/ Effects of net energy levels on feed intake pattern and growth performance of S0003356100007182. growing-finishing gilts monitored using a computerized feed intake recording Young, J.M., Cai, W., Dekkers, J.C.M., 2011. Effect of selection for residual feed intake on system. J. Anim. Sci. 97, 175. feeding behavior and daily feed intake patterns in yorkshire swine. J. Anim. Sci. 89, 639–647. https://doi.org/10.2527/jas.2010-2892. 17