Feed Reward Research Articles

Learning, health and productivity of group-housed sows conditioned to signal-feeding under realistic husbandry conditions

Signal-feeding is a discrimination task for pigs where individual acoustic signals indicate the availability of feed at an electronic feeding station. It has been shown that this feeding technique has the potential to reduce agonistic interactions between pigs and improve animal welfare. As a side effect, signal-feeding was found to facilitate the accessibility of the feeding stations. Therefore, this feeding system may have the potential to improve the pigs’ overall feed supply and productivity. Signal-feeding has been successfully adapted for use with adult sows at conventional electronic feeding stations and was tested under near-practical conditions. The present field study was conducted to test the applicability of signal-feeding under the realistic conditions of a conventional piglet rearing facility. The aim was to verify that sows can cope with the signal discrimination task under the limitations of intensive livestock farming which comprises large groups with high feed concurrence and only one training opportunity per day. Furthermore, the study should confirm any positive effects on the animals’ health and welfare and, for the first time, substantiate the economic viability of signal-feeding by demonstrating a reduced workload and increased productivity compared to conventional electronic feeding stations.As a main result of the study, we observed that a single training opportunity per day and training in groups of more than 30 sows resulted in a training duration that ranged from 12 to 38 days for the majority of sows. The liquid feeding system utilised was found to be inadequate for reward delivery. It left feed remains on the station floor and could not provide the feed reward in sufficiently close temporal relation to the animal behaviour. The feed remains in combination with the somewhat random reward delivery created sufficient motivation for some sows to continue visiting the feeding station. As a result, no improvements in the accessibility of the feeder were observed. Under these conditions, we did not find any effects on the productivity of the sows in terms of the number of piglets born alive or the number of weak or crushed piglets. However, the training itself led to an increased number of shallow skin lesions and higher medical costs at the beginning of the study when the majority of sows received their conditioning. This effect disappeared later, when only replacement sows were trained.This study demonstrates that in sows an automated operant conditioning of a complex signal discrimination task is possible even under the considerable limitations of intensive livestock farming. However, future studies should utilise a more direct and better accessible reward system to reduce the training duration and improve the conditioning of the sows. Only then, will it be possible to accurately assess the animal welfare aspects and economic potential of signal-feeding.

Investigation of donkeys learning capabilities through an operant conditioning

Despite donkeys being involved in various activities with humans, their cognitive and learning abilities are still little known. A deeper understanding of their perceptive, cognitive and learning processes is, thus, necessary to preserve their well-being and establish a good human-animal bond. An operant conditioning task was applied to explore donkeys’ learning abilities. Nine out of 14 adult and non-working donkeys of both sexes fully completed a three-phases training procedure. In the first phase, animals could approach the manipulandum, a specifically designed cabin with a button on the front side. Thereafter, donkeys had to learn how to interact with the button to obtain a feed reward. To evaluate donkeys’ learning capabilities, two linear models were built: i) a fixed-effects model, exploring how the average between-pressures time (BPT) was affected by individual characteristics (i.e. age, sex, and donkeys’ height at the withers) and the training session; and ii) a mixed-effects model to evaluate the difference in the average between-pressures time among consecutive sessions (BPTcs) in the function of the animals’ characteristics, including the sessions-lags as the random effect. In the first model, all the explanatory variables resulted significantly associated with the BPT observed variability. Male donkeys presented a BPT significantly higher, increased by 23.14 s (S.E.=9.71, p = 0.003) compared to females of the same height and age. Age was significant with a positive coefficient (Est.=1.21, S.E.=0.55, p = 0.032). The ‘high’ height class estimate was significant (Est.=13.06, S.E.=6.26, p = 0.032), while no significant effects were identified between the ‘medium’ and ‘short’ and the ‘medium’ and ‘high’ height classes. Lastly, the variable ‘session’ was significant with a negative coefficient (Est.=−10.64, S.E.=1.56, p < 0.0001), indicating an increase in the average speed to perform the desired behaviour for each additional training session. In the second model, the variable ‘sex’ was the only predictor significantly associated with the BPTcs, indicating that the male group progressively improved performance time faster than females (Est.=−8.71, S.E.=4.19. p = 0.045). This pilot study: i) provides insights into donkeys learning abilities by applying an effective methodology for operant conditioning; ii) it highlights how intrinsic animal characteristics might affect asses training performances, although further points need to be explored in future research; iii) it confirms that feed represents an effective positive reinforcement in operant conditioning with donkeys. The development of appropriate handling and training methods, respectful of animals’ subjective experiences and based on positive practices, can improve donkeys’ welfare and their relationship with humans.

Impact of environmental complexity and stocking density on affective states of rainbow trout (Oncorhynchus mykiss).

Environmental condition, such as environmental complexity or stocking density, can directly or indirectly influence animal emotion and ultimately, affective state. Affective states of animals can be assessed through judgement bias tests, evaluating responses to ambiguous situations. In this study, we aimed to determine whether environmental complexity and stocking density impacted rainbow trout affective state. Rainbow trout (n = 108) were housed in recirculating aquaculture systems under commercial conditions while trained at tank-level to discriminate between a positively reinforced chamber (feed) in one location and a negative chamber (positive punishment; chase by net for 1s) in the opposing location. Fish from successful tanks (two out of five tanks) were then housed in treatment tanks of either high- or low- environmental complexity at either high (165 fish/m3) or low (69 fish/m3) stocking density. Trained fish were tested for latencies to approach three intermediate, ambiguous chambers. Fish housed in high-density tanks were faster to enter all chambers than those housed in low-density tanks (8.5s vs. 15.2s; P = 0.001), with faster entries into the positive (7.4s vs. 15.2s; P = 0.02) and near-negative chambers (10.2s vs. 17.4s; P = 0.006), suggesting that these fish were more optimistic to receive a feed reward. Tank complexity did not affect test outcomes. No differences between treatments were observed between body weight, length, and plasma cortisol. Overall, rainbow trout are capable of discriminating between cues during a judgement bias test and fish housed in high-density environments respond more optimistically in ambiguous situations compared to fish in low-density environments.

Movement or more food? A comparison of motivation for exercise and food in stall-housed sows and gilts

Despite being phased out in the EU, several states in the USA and other select countries such as Australia, gestation stalls remains a common sow housing system worldwide. In Canada, the 2014 Code of Practice for the Care and Handling of Pigs permits the operation of existing stall barns, provided that bred female pigs are given opportunities for a greater freedom of movement, such as access to periodic exercise. How stall-housed sows perceive being confined and whether they are motivated for a greater freedom of movement has not been investigated. The objectives of this study were to determine the motivation of bred female pigs with different prior stall experience (PSE) to receive access to an alleyway for three minutes (Movement), and to compare this to their motivation for a feed reward (Food: 0.2 kg per reward). Stall-housed gilts (n = 12) and sows (n = 12, parity 2–4) were trained to use an operant panel with two buttons: an active button (AB – push counts resulted in a reward of feed or movement), and a dummy button (DB – push counts not rewarded). Animals were tested on an ascending fixed ratio (FR), with the required number of AB presses increasing by 50 % each consecutive day. Upon reaching their individual maximal FR for one reward, animals were retrained and tested on the opposite button and received the alternate reward. The maximum AB push count achieved during a testing session, the highest price paid: HPP, was used as a measure of sow motivation for the reward. Animal behaviour when out of the stall and the percentage of AB presses were also recorded. The HPP was influenced by PSE and reward type. Sows showed a greater HPP for feed than movement, but for gilts it did not differ. Sows had a greater motivation to access feed than gilts, however, gilts and sows did not differ in their motivation to access movement (mean HPP ± SEM, sows Food: 369.25 ± 56.47 vs Movement: 68.5 ± 13.61; gilts Food: 211.67 ± 47.73 vs Movement: 77.75 ± 19.47, F(1, 44) = 4.23, P ≤ 0.05). Results suggest that stall-housed gestating sows and gilts are similarly motivated to access time out of their stall. The presence of the motivation for movement is important for sows even though they had a greater motivation to access a feed reward, and the motivation for a feed reward increases with PSE.

Dairy cows fed a low energy diet before dry-off show signs of hunger despite ad libitum access

Drying-off is one important management step in commercial dairy farms and consists of ceasing milk production artificially at a specific point in time, generally 2 months before the next calving. Drying-off typically comprises dietary changes as well as gradual or abrupt changes in daily milking frequency, which may challenge the welfare of high-yielding cows. This study investigated the isolated and combined effects of different feed energy densities (normal lactation diet versus energy-reduced diet, both offered ad libitum) and daily milking frequencies (twice versus once) on the feeding motivation of dairy cows on two separate days prior to dry-off (i.e. the day of last milking) using a push-gate feeder. During both days, cows on the energy-reduced diet pushed more than five times more weight to earn the final feed reward and were nearly ten times faster to feed on the first reward than cows on the normal lactation diet. Illustrating the importance of developing more animal welfare-friendly dry-off management, these results illustrate that cows show signs of hunger prior to dry-off when provided a diet with reduced energy density, although offered for ad libitum intake.

Positive or negative emotion induced by feeding success or failure can affect behaviors, heart rate and immunity of suckling calves

This study investigated behavioral and physiological responses of suckling calves to either positive or negative situations. Twelve healthy female suckling Holstein calves at approximately 30 days of age were randomly selected and allocated to two groups: a) a positive situation (POS) group that received a feed reward and b) a negative situation (NEG) group that experienced food frustration. Behavior, heart rate (HR), and immune parameters were measured during the emotion-inducing periods and analyzed via ANOVA. The results indicated that the POS calves showed significantly increased locomotive play but decreased self-grooming, head shaking, and tongue-rolling behaviors as compared to the NEG counterparts (p < .05). The NEG calves had the highest HR (p < .01) after negative emotion induction. Salivary cortisol (S-CORT) level significantly increased following both positive and negative emotional induction (p < .05). In humoral immunity, serum concentrations of IgA were not significantly affected by either the positive or negative situation (p > .05), but the salivary SIgA level of the POS calves was higher than that of the NEG group (p < .01). Serum IL-2 and IL-3 levels in the POS calves were significantly higher than those of the NEG group (p < .01). However, the serum TNF-α level in the POS group was significantly lower than in the NEG group (p < .05). The results suggest that hunger satiation or food frustration may induce different behavioral and physiological responses in suckling calves, presenting an interesting set of reactions appropriate in response to the emotional situations elicited by this short-term stimulus

The effect of herd mentality on dairy heifers conditioned to traffic through audio cues

The objective for the present trial was to understand whether dairy heifers could be trained to respond to an audio cue paired with a feed reward. The use of acoustic conditioning to induce cattle movement has not previously been tested with animal-mounted devices to call cattle both individually and as a group. Five heifers underwent testing for 6 days as part of an 18-day field trial (12 days of conditioning). The 6-day testing and data-collection period involved the heifers being called via a smartphone device mounted on the cheek strap of a halter. Heifers were called either as individuals or as a group. When the audio cue was sent, heifers were expected to traffic from a group-holding area to a feeding area (~80-m distance) to receive an allocation of a grain-based concentrate. Heifers were significantly (P = 0.001) more likely to approach the feeding area when called as a group (91% response rate) than when they were called as individuals (67% response rate). When heifers did respond to being called, their time to traffic to the feed area was quicker (P &lt; 0.001) when they were called as a group (77.9 ± 55.4 s) than when they were called as individuals (139.3 ± 89.2 s). The present trial has shown that animals can be trained to respond to an audio cue paired to a feed reward, highlighting the potential for acoustic conditioning to improve voluntary cow movement with an animal-mounted device. It also highlights the limitations of cattle responding to being called individually compared with being called as a group.

1199 Analysis and review of publicly available GreenFeed results

The GreenFeed is a novel tool for measurement of enteric methane (CH4) from ruminant animals. The GreenFeed method uses repeated short-term measurement of CH4 emitted from the animal in a feed trough while animals are receiving a feed reward. To date, over 50 papers, conference proceeding, and reports are available that have used GreenFeed. The research covers a wide range of applications including method comparisons, animal genetics and dietary studies, and on-farm use. In many studies, variability and repeatability of the GreenFeed data are reported. The objective of this study was to aggregate the public data, analyze the results for overall accuracy compared to reference methods, summarize variability and repeatability of CH4 emissions across applications, and to determine the strength of individual animal ranking relationships. Overall, 22 method comparison trails ranging in herd averaged emission from 150–485 g/d have been completed comparing GreenFeed to either the Sulfur Hexafluoride tracer method (SF6, 11 trials), respiration chambers (9 trials), or model predictions (2 trials). The herd averaged results for the reference method compared to GreenFeed showed no significant slope bias (Reference = 0.99 × greenFeed, R2 = 0.99) and the average absolute mean error for all trails was 6.4%. For individual animal CH4 emissions, GreenFeed measured CH4 emissions were positively and moderately-to highly correlated with respiration chamber, SF6, or modeled CH4 in all but 4 trials. In the 4 trials with less agreement for individual animals, the number of GreenFeed samples per animal was very low for some animals (4–10 samples) or there was no chamber replication. GreenFeed measured CH4 for individual animals also showed moderate correlations with DMI in three studies (r2 = 0.73, r2 = 0.61, r = 0.77), and in other studies was found to be higher (r2 = 0.42 and r2 = 0.47) than for SF6 (r2 = 0.17, r2 = 0.08). Between animal variation in GreenFeed CH4 emissions were found to be the same, or lower than between animal variation in DMI, BW, or Respiration Chambers, or SF6 in all but one trail. Long term repeatability (R) of GreenFeed CH4 emissions on forage based diets was R = 0.70- 0.88 in 7 different trials. Overall, the GreenFeed method produced similar absolute CH4emissions estimates to reference methods, demonstrated the ability to rank animals, produced low but accurate between animal variability, and was highly repeatable on forage diets.

Sprinkler flow rate affects dairy cattle avoidance of spray to the head, but not overall, in an aversion race

Spraying cattle with water elicits behavioral responses ranging from willing use to avoidance of wetting the head or entire body. This variation may be partly explained by sprinkler flow rate, as higher flow rates result in greater heat abatement and spray impact. An aversion race can be used to evaluate the emotional valence cows associate with treatments, based on the prediction that they will approach negative (vs. positive) stimuli more slowly and require greater pressure from a handler. Our objective was to evaluate how heat load affects the degree to which dairy cattle show reluctance to approach sprinklers with different flow rates in an aversion race. We compared unsprayed treatments (n=8 with feed, n=16 without) to 0.4L/min (n=14) and 4.5L/min (n=17) sprinklers, which differed eight-fold in spray impact (1.1 vs. 8.9kPa, respectively). Over 10 trials, cows were required to approach treatments in a covered, narrow raceway (average outdoor air temperature, mean±SD, 31±3.8°C). We compared time to complete the race (transit time), the pressure handlers applied, and head posture. Overall, handlers applied pressure half as often when feed was offered but there were no other treatment differences in handling pressure or transit time. Thus, we excluded the feed treatment when examining interactions with heat load and sprinkler flow rate. In warmer weather, transit time increased (by 13s per 10°C increase). As their respiration rate increased, unsprayed cows moved more slowly (by 7s per 10 breaths/min increase), whereas sprayed cows did not. Handling pressure was not affected by heat load or spray application, but head posture depended on sprinkler flow rate. When cows approached 4.5L/min, they lowered their heads nearly five times as often compared to 0.4L/min or no spray (proportion of trials±SE, 0.81±0.05, 0.17±0.06, and 0.15±0.05, respectively). This may have been an attempt to reduce exposure of sensitive body parts to higher-impact spray. Indeed, when we applied von Frey monofilaments to the ear and shoulder (n=56), the ear had a lower response threshold (greater sensitivity) than the shoulder [mean (95% confidence interval), 0.2 (0.1 to 0.4) vs. 1.2 (0.5 to 2.5) g of force, respectively, back transformed from natural log values]. In conclusion, each measure provided different insights about reluctance to approach treatments. Transit time increased in response to heat load rather than aversion. Handling pressure reflected willingness, as it was needed less often when cows were offered a feed reward, but this measure did not differ among sprinkler treatments. Therefore, although cows lowered their heads to avoid exposing sensitive areas to 4.5L/min spray, they did not show overall aversion to higher-impact spray.

Effect of locomotion score on sows’ performances in a feed reward collection test

Sows housed in groups have to move through their pen to fulfil their behavioural and physiological needs such as feeding and resting. In addition to causing pain and discomfort, lameness may restrict the ability of sows to fulfil such needs. The aim of our study was to investigate the extent to which the mobility of sows is affected by different degrees of lameness. Mobility was measured as the sow’s willingness or capability to cover distances. Feed-restricted hybrid sows with different gait scores were subjected to a feed reward collection test in which they had to walk distances to obtain subsequent rewards. In all, 29 group-housed sows at similar gestation stage (day 96.6±7 s.d.) were visually recorded for gait and classified as non-lame, mildly lame, moderately lame or severely lame. All sows received 2.6 kg of standard commercial gestation feed per day. The test arena consisted of two feeding locations separated from each other by a Y-shaped middle barrier. Feed rewards were presented at the two feeders in turn, using both light and sound cues to signal the availability of a new feed reward. Sows were individually trained during 5 non-consecutive days for 10 min/day with increasing barrier length (range: 0 to 3.5 m) each day. After training, sows were individually tested once per day on 3 non-consecutive days with the maximum barrier length such that they had to cover 9.3 m to walk from one feeder to the other. The outcome variable was the number of rewards collected in a 15-min time span. Non-lame and mildly lame sows obtained more rewards than moderately lame and severely lame sows (P<0.01). However, no significant difference was found between non-lame and mildly lame sows (P=0.69), nor between moderately lame and severely lame sows (P=1.00). This feed reward collection test indicates that both moderately lame and severely lame sows are limited in their combined ability and willingness to walk, but did not reveal an effect of mild lameness on mobility. These findings suggest that moderately and more severely lame sows, but not mildly lame sows, might suffer from reduced access to valuable resources in group housing systems.

Measuring motivation for appetitive behaviour: food-restricted broiler breeder chickens cross a water barrier to forage in an area of wood shavings without food.

Broiler breeders (parents of meat chickens) are selected for fast growth and become obese if fed ad libitum. To avoid this and maintain good health and reproductive ability, they are feed restricted to about 1/3 of what they would eat ad libitum. As a result, they experience chronic hunger and exhibit abnormal behaviour patterns that may indicate stress and frustration. One approach to measuring hunger is to observe how much birds will work, such as pecking a key, for access to more or different types of food. However, the sight, smell, and feedback from consumption of the feed reward changes the context and may artificially raise feeding motivation. To avoid this, we tested broiler breeders in an apparatus in which they could work for access to a wooden platform covered in wood shavings by crossing a water runway which increased in length and depth in 8 successive tests. In the wood shavings area, they could perform exploratory and foraging behaviour (the appetitive phase of feeding) but were never rewarded with feed. Sixty birds were divided into three feed quantity treatments: commercial restriction (R), and twice (2R) or three times (3R) this amount. Overall, birds fed R worked harder to reach the wood shavings area (reached it in a larger number of tests) than 2R and 3R birds (P<0.001). More restricted birds took less time to reach the area (P<0.001, R<2R<3R) and spent more time foraging while there (P<0.001, R>2R>3R). This indicates that restricted-fed birds were hungry and willing to work for the opportunity to forage even though food was never provided, suggesting that their motivation to perform the appetitive component of feeding behaviour (foraging/food searching) was sufficient to sustain their response. Thus food restriction in broiler breeders is a welfare concern. However these methods could be used to test alternative feeding regimes to attempt to find ways of alleviating hunger while still maintaining healthy growth and reproduction in these birds.

Influence of provision of concentrate at milking on voluntary cow traffic in a pasture-based automatic milking system

The success of an automatic milking system is generally reliant upon the voluntary movement of cows around the farm system and the correct management of incentives to achieve a targeted level of cow traffic. The present study investigated the effect of providing a small feed reward as an incentive at milking on the premilking voluntary waiting time of cows milked on a prototype robotic rotary in an Australian pasture-based dairy. The 2 treatments were “feed on” (concentrate offered at milking) and “feed off” (no concentrate offered at milking), with data from a single herd of 168 lactating dairy cows collected over 16d. A survival analysis with time-varying covariates was used to model the voluntary waiting times of cows in the premilking yard. The median time cows spent waiting before milking was 129min and after 4h just over 70% of the cows had exited the yard (volunteered for milking). When feed was provided, cows were faster to exit the premilking yard (shorter time spent waiting) and waited just over half the time (0.53×) they did during the “feed off” treatment. Heifers exited the premilking yard more rapidly than cows in later lactations, with older cows spending at least 1.40 times longer in the yard before milking. Average daily milk yield along with stage of lactation and fetching cows from the paddock also influenced cow traffic in the premilking yard. As the number of cows in the premilking yard increased, voluntary waiting time also increased. At a queue length of 20 or more cows, the negative effect on waiting time of an additional cow entering the yard was less than that when fewer than 20 cows were present. Results demonstrated that feeding a small reward on the robotic rotary platform can reduce the time cows spend in the premilking yard, leading to a potential reduction in the risk of congestion at the dairy, particularly during times of high demand. Minimizing congestion will likely benefit multiple aspects of the voluntary milking operation, including a potential improvement in robot utilization, a reduction in unnecessary time spent off pasture by cows in the milking herd, promoting cow welfare through reducing the risk of lameness, and enhancing productivity. Targeting strategies to minimize queue length to less than the threshold length, which in this study was 20 cows, could result in reduced time spent in the premilking yard.

35 LEARNING AND MEMORY BEHAVIORAL PATTERNS IN CLONED DOGS

Recently, in order to advance biological technology and increase the number of elite dogs that possess unique abilities, researchers have put more focus on cloning superior dogs that have been acknowledged in their respective fields. These experiments depend on the known fact that cloned dogs will be both physically and psychologically similar to those that provided the somatic cells. However, little research has focused on whether it is genetic or posteriori factors that influence abilities to accomplish tasks. In this experiment, cloned beagles that have been produced from one somatic cell and thus have the same genetic information were tested on learning and memory behaviours. This experiment was performed to investigate the similarity in behavioural patterns between these cloned individuals (n = 6). A Y-maze test, which is commonly applied for evaluating learning and memory, was performed using 12 Beagles, 6 of which were cloned dogs and the other 6 were naturally bred controls (n = 6). One snack was placed at the end of each arm of the Y-maze. The snack that was placed in the east arm was accessible to the dogs, whereas the other was blocked using a plastic fence. All 12 dogs were trained before the experiment, where they were sent through the maze 10 times and allowed to obtain the snack. Following this training period, the dogs were retested 3 times to assess learning and working memory. The first trial (Day 0) was performed 1 day after the training period, while the second (Day 7) and third (Day 14) trials were performed at 1-week intervals. In each trial, the dog was given 60 s to make a choice between the east or west arm of the Y-maze. If correct, the dog received a feed reward. After consuming the reward, the dog was picked up by the experimenter and placed outside in preparation for the next dog. In every trial, each dog was sent through the Y-maze 5 times. Using the Harvard Panlab software and a live video image, the latency to choice was measured. All cloned dogs reached the performance criterion with 100% correction on Day 0 of acquisition. However, all naturally bred dogs reached the performance criterion (85% correction) at a lower rate than those of cloned dogs. Correct choices were maintained in all experimental dogs during re-assessment time. Mean latency to choice showed no significant differences between naturally bred controls and cloned dogs. Interestingly, when the standard deviation (s.d.) of the latency to reach the target was compared, the s.d. within the cloned group was significantly lower than that within the control group only on the first assessment (Day 0). However, no significant differences were shown during the second (Day 7) and third (Day 14) trials. Therefore, it appears that genetically identical cloned dogs do not show greater consistency in their learning and memory behaviour than litters of naturally bred control dogs. The learning and memory ability of cloned dogs were not different from those of naturally bred control dogs. This research was supported by the SNU Undergraduate Research Program.

Performance of sheep in a spatial maze is impeded by negative stimuli

Exposure to negative stimuli or stress can manifest in general changes in cognitive processing. This study aimed to investigate if a spatial maze task could be used to identify stress-induced differences in the cognitive performance of sheep. Two negative stimuli were used to test the hypothesis. For a negative pre-treatment (‘dog’ pre-treatment), sheep were moved individually to a holding yard at the beginning of the maze where they were exposed to a dog for 3min, for 5 consecutive days. Alternative to the dog pre-treatment, sheep were moved in small groups to the same holding yard, for the same amount of time, where they received a feed reward (‘food’ pre-treatment). For a during-test stimulus, white noise was played as sheep moved through the maze (‘noise’ treatment). Sixty-four male castrated lambs were allocated to one of four groups: dog and noise, food and noise, dog and no noise, or food and no noise. Sheep traversed the maze on 3 consecutive days and the total time to complete the maze, the number and the duration of errors made were used to assess cognitive performance. Maze results were analysed using GLMM, LMM and linear contrasts. The noise increased both total time (140s vs. 105s, P=0.043) and error time (67s vs. 56s, P=0.044) on day 1. The dog pre-treatment increased error time compared to the food pre-treatment (81s vs. 41s, P=0.041) and tended to increase the number of errors made on day 1 (1.5 errors vs. 1.2 errors, P=0.057). Neither noise nor dog pre-treatment influenced cognitive performance on days 2 or 3. Results suggest that both stimuli affected cognitive performance in the maze by impeding initial problem solving. The maze used demonstrates the ability to identify differences in cognition.

Acoustic-reward learning as a method to reduce the incidence of aggressive and abnormal behaviours among newly mixed piglets

The aim of the study was to test whether aggressive actions among piglets could be redirected by an automatically generated sound signal followed by a sweet food reward. Per round, four litters of 25-day-old piglets (BHZP breed) were trained 5 times per day over 8 days to expect a sweet feed reward from a dog feeder after hearing a specific sound. In total 144 piglets in 14 entire litters were trained in five trials. At the end of the training 71% of the piglets were around the feeder 5 s after the feeder sound. After the training period, the piglets were weaned and mixed in two pens, 12 piglets per pen. During 2 days (3 h/day) after mixing two observers (one per pen) hidden behind a wooden wall activated the feeder when aggressive or abnormal behaviour started. A total of 616 aggressive events and 31 incidences of abnormal behaviour (ear biting) were used for the analysis. The logistic regression showed that the type of behaviour had a significant effect on the piglets’ response to the feeder sound (P &lt; 0.001). The results showed the possibility of interruption of the aggressive behaviours such as head thrust [odds ratio (OR) = 0.43], jump on other (OR = 0.56) or attack with bite (OR = 0.61). Ear biting was very unlikely to continue (OR = 0.55). The risk of continuing elevated aggression level behaviours was doubled in the event of chasing (OR = 2.16) and the risk that fight would continue after the feeder sound was released was 7 times higher (OR = 7.89). Categorical analysis showed a significant effect (&lt;0.001) of the time intervals t ≤ 1 s and 1 s &lt; t ≤ 3 s on interruption of aggression by the feeder sound release. The piglets’ response to the feeder sound differed significantly between the experimental days (P &lt; 0.001). On the second day of mixing, the feeder sound interrupted 74.9% of aggressive events, compared with 33.7% on the first day. The results suggest that acoustic-reward treatment can distract pigs from performing certain aggressive behaviours and ear biting in piglets when properly applied in time.

Qualitative behavioural assessment of the motivation for feed in sheep in response to altered body condition score

Qualitative behavioural assessment (QBA) has been used to quantify the expressive behaviour of animals, and operant tests have been used to quantify measures of behavioural need. In this study we compared measures of behavioural expression and behaviour in operant tests. We examined the behavioural expression of pregnant ewes of body condition score (BCS) 2 and 3. The ewes were exposed to a feed motivation test in which they received a food reward. Pregnant ewes (48–70 days gestation) were assessed during a food motivation test after they had been maintained at BCS 3 (n = 7) or given a decreasing plane of nutrition that resulted in slow loss of 1 BCS unit (over 10–12 weeks; n = 7) or a fast loss of 1 BCS unit (over 4–6 weeks; n = 7). The feed motivation test involved ewes having the opportunity to approach a food reward and then being moved a given distance away from the reward by an automatic gate; they could then subsequently return to the feeder. Continuous video footage of each ewe during one cycle of the gate (approaching and returning from the food reward) was shown in random order to 11 observers who used their own descriptive terms (free-choice profiling methodology; FCP) to score the animals using QBA. Data of the assessment were analysed with generalised Procrustes analysis (GPA), a multivariate statistical technique associated with FCP. The research group also quantified the feeding behaviour of sheep in the same clips. These behaviours included how sheep approached the feeder, behaviours exhibited at the feeder, and how sheep returned from the feeder. There was consensus amongst observers in terms of their assessment of behavioural expression of the sheep (P &lt; 0.001). The GPA found three main dimensions of assessed behavioural expression in the sheep, which together explained 44% of the variation observed. GPA dimension 1 differed between the three treatment groups (P &lt; 0.05): ewes maintained at BCS 3 scored low on GPA dimension 1 (i.e. were described as more calm/bored/comfortable) compared with ewes that had a slow declining BCS (described as more interested/anxious/excited). GPA dimension 2 scores were not significantly different between treatment groups. However, quantitative behaviours exhibited by sheep during the clips were correlated with qualitative behavioural assessments made by the observers. Animals that spent more time ‘sniffing and looking for more feed’ were attributed lower GPA 2 scores (described as more hungry/searching/excited) (P &lt; 0.05), and animals that ‘did not walk directly to the food reward (but stopped along the way)’ were attributed significantly higher GPA 2 scores (more curious/intimidated/uneasy) (P &lt; 0.01). GPA dimension 3 scores also did not differ between the treatment groups; however, sheep that had a higher number of feeding events during the entire 23-h feed motivation test were attributed lower GPA dimension 3 scores (they were described as more hungry/bold/interested) (P &lt; 0.05), and sheep that consumed a larger amount of the feed reward were attributed higher GPA dimension 3 scores (more curious/concerned/reserved) (P &lt; 0.05). We conclude that QBA is a valuable method of assessing sheep behavioural expression under the conditions tested, in that it provided an integrative characterisation of sheep behavioural expression that was in agreement with quantitative behavioural measures of feeding.

Cognitive Enrichment in Piglet Rearing: An Approach to Enhance Animal Welfare and to Reduce Aggressive Behaviour

It is known that pigs raised in enriched environments express less aggressive behaviour. For this reason, a new method of cognitive environmental enrichment was experimented at the University of Veterinary Medicine Hannover, Germany. In the first phase, 78 suckling piglets were trained to learn the link between a sound given by an electronic feeder and a feed reward in the form of chocolate candies during a period of 8 days. In the second phase, the same piglets were used in resident-intruder tests to verify the potential of the feeding system to interrupt aggressive behaviour. The analysis of all training rounds revealed that piglets learned the commands during 8 days of training and the interest of the piglets increased within training days (P < 0.05). In the resident-intruder test, 79.5% of aggressive interactions were broken by feeder activation. In interactions where either the aggressor or the receiver reacted, a high number of fights were stopped (96.7% versus 93.1%) indicating that it was not relevant if the aggressor or the receiver responded to the feeder activation. We conclude that the electronic feeding system has the potential to be used as cognitive enrichment for piglets, being suitable for reducing aggressive behaviour in resident-intruder situations.

Disrupting motivational sequences in chicks: Are there affective consequences?

The ‘reward cycle’ conceptualises reward acquisition as a cyclic phenomenon divided into three motivational stages with related emotional or affective states. For feeding behaviour such a cycle consists of an appetitive stage characterised by foraging and exploration linked to emotions such as wanting and anticipation, a consummatory stage with eating behaviour linked to liking and pleasure, and a post-consummatory stage linked to satiety and relaxation with behaviour like resting and preening. In this study we investigated whether disturbing the feed reward cycle in laying hen chicks, by denying access to parts of a pen designed to accommodate the stages of the cycle (litter area ‘appetitive’; feed area ‘consummatory’; perches and dark area ‘post-consummatory’), resulted in a more negative affective state. To test this, we used a spatial cognitive bias task in which a bowl in one location in the test arena was associated with a positive outcome (mealworm), and in a different location with a negative outcome (unpalatable puffed rice soaked in quinine sulphate). Three ambiguous probe locations were presented during the test. Chicks (n=22) discriminated between the positive and negative location as evidenced by a significant difference in times to reach these locations (mean difference variable-feed treatment 22.1±8.8s; closed-litter treatment 23.3±6.5s; closed-dark treatment 24.4±4.9s and baseline mean difference 22.3±6.4s). Chicks denied access to the litter area was significantly quicker to reach the probe near the negative location than when denied access to the feed area (mean 8.9±1.7 vs. 18.6±1.7) – an ‘optimistic’ judgement of ambiguity indicative of a less negative affective state when denied litter compared to when denied feed. Relative to the initial baseline cognitive bias tests, all treatments resulted in slower times to reach the negative location (closed-dark: 14.9±1.9; variable-feed: 12.6±1.9; closed-litter: 13.7±1.9) and shorter times to the positive location (closed-dark: −7.3±1.7; variable-feed: −7.2±1.7; closed-litter: −7.3±1.7). Continuing improvement in learning of the positive versus negative location discrimination following baseline tests, or a change in perception of the incentive value of the positive and negative outcomes, may explain this finding. There was no evidence that variations in fearfulness or sociality (measured in tonic immobility and social reinstatement tests) affected the outcome of the cognitive bias tests. There seems to be different reactions to disrupting different parts of a reward cycle and further investigations into the links between affect and motivational sequences may provide a better understanding of the affective importance of different resources.

Prefrontal cortex activity, sympatho-vagal reaction and behaviour distinguish between situations of feed reward and frustration in dwarf goats

Recent concepts relating to animal welfare accept that animals experience affective states. These are notoriously difficult to measure in non-verbal species, but it is generally agreed that emotional reactions consist of well-coordinated reactions in behaviour, autonomic and brain activation. The aim of the study was to evaluate whether each or a combination of these aspects can differentiate between situations presumed to differ in emotional content. To this end, we repeatedly confronted dwarf goats at short intervals with a covered and an uncovered feed bowl (i.e. presumably frustrating and rewarding situations respectively) whilst simultaneously observing their behaviour, measuring heart-rate and heart-rate variability and haemodynamic changes in the prefrontal cortex using functional near-infrared spectroscopy. When faced with a covered feed bowl, goats occupied themselves at locations away from the bowl and showed increased locomotion, while there was a general increase in prefrontal cortical activity. There was little indication of autonomic changes. In contrast, when feed was accessible, the goats reduced locomotion, focused their behaviour on the feed bowl, showed signs of sympathetically mediated arousal reflecting anticipation and, if any cortical activity at all was present, it was concentrated to the left hemisphere. We thus observed patterns in behaviour, sympathetic reaction and brain activity that distinguished between a situation of frustration and one of reward in dwarf goats. These patterns consisted of a well-coordinated set of reactions appropriate in respect of the emotional content of the stimuli used.

Does reduction of fearfulness tend to reduce pessimistic-like judgment in lambs?

Recent studies emphasize the role of interactions between emotions and judgment of environment to better assess affective state in animals. Diazepam offers a way to pharmacologically manipulate the affective state. This drug is generally used to reduce negative affective states, mainly by reducing fearfulness. Here we investigated whether a reduction of fearfulness through a pharmacological approach could reduce pessimistic-like judgment in lambs. We tested the effects of diazepam (0.10mg/kg), a benzodiazepine known for its anxiolytic properties in many species, in 5-months old female lambs submitted to two experiments: the first one to assess their fear-related reactions (cortisol, cardiac and behavioural measures) to isolation and suddenness tests, and the second one to assess their putative judgment bias. In experiment 1, the lambs treated with diazepam were less frightened by isolation and suddenness than the control lambs injected with vehicle, which tended to confirm the anxiolytic effect of diazepam on sheep. In experiment 2, the lambs were first trained to perform a spatial location task to test for judgment bias in sheep. This task required a go/no-go response according to the location of a bucket in a pen, with one location being reinforced positively (with a feed reward) and the other negatively (with a fan-forced blower). Once trained, lambs (n=16) were exposed twice (10min and 3h after injections) to five consecutive bucket locations: the two learnt locations plus three unrewarded, ambiguous locations set between the learnt locations. Control lambs increased their latency to approach one of the ambiguous bucket locations (P<0.05), while treated lambs maintained the same latency to approach this location. This difference may reflect a bias in judgment in relation to fearfulness: treated lambs seemed to display a more positive judgment of an ambiguous event than control lambs. Reduction of fearfulness may thus induce a more positive affective state but this requires further investigation, with additional studies comparing the efficacy of the cognitive bias approach to other measures of affective state.