Gerhard Manteuffel

Assessment of positive emotions in animals to improve their welfare

It is now widely accepted that good welfare is not simply the absence of negative experiences, but rather is primarily the presence of positive experiences such as pleasure. However scientific investigation of positive emotions has long been neglected. This paper addresses two main issues: first, it reviews the current state of scientific knowledge that supports the existence of positive affective states in animals and, second, it suggests possible applications of this knowledge that may enhance quality of life under animal management conditions. In the first part of the paper, recent advances in psychology and neuroscience are reviewed to provide pragmatic frameworks based on cognitive processes (such as positive anticipation, contrast and controllability) for further investigations of positive emotions in animals. Thereafter, the neurobiological bases of positive emotions are highlighted in order to identify behavioral and physiological expressions of positive experiences in animals. Monitoring both the autonomic nervous system (via heart rate and its variability) and the immune system could offer relevant tools to better assess emotional states in animals, complementary to classical adrenocortical measures. In the second part of the paper, useful strategies for enhancing positive experiences (such as physical, social and cognitive enrichment or putative genetic selection) are outlined. Then this paper emphasizes practical applications for assessing and promoting positive emotions that may help in providing animals with a better quality of life. Play, affiliative behaviors and some vocalizations appear to be the most promising convenient indicators for assessing positive experiences in laboratory and farm animals under commercial conditions.

Exploration of the hypothalamic–pituitary–adrenal function as a tool to evaluate animal welfare

Measuring HPA axis activity is the standard approach to the study of stress and welfare in farm animals. Although the reference technique is the use of blood plasma to measure glucocorticoid hormones (cortisol or corticosterone), several alternative methods such as the measurement of corticosteroids in saliva, urine or faeces have been developed to overcome the stress induced by blood sampling itself. In chronic stress situations, as is frequently the case in studies about farm animal welfare, hormonal secretions are usually unchanged but dynamic testing allows the demonstration of functional changes at several levels of the system, including the sensitization of the adrenal cortex to ACTH and the resistance of the axis to feedback inhibition by corticosteroids (dexamethasone suppression test). Beyond these procedural aspects, the main pitfall in the use of HPA axis activity is in the interpretation of experimental data. The large variability of the system has to be taken into consideration, since corticosteroid hormone secretion is usually pulsatile, follows diurnal and seasonal rhythms, is influenced by feed intake and environmental factors such as temperature and humidity, age and physiological state, just to cite the main sources of variation. The corresponding changes reflect the important role of glucocorticoid hormones in a number of basic physiological processes such as energy metabolism and central nervous system functioning. Furthermore, large differences have been found across species, breeds and individuals, which reflect the contribution of genetic factors and environmental influences, especially during development, in HPA axis functioning. Usually, these results will be integrated with data from behavioral observation, production and pathology records in a comprehensive approach of farm animal welfare.

Visual discrimination learning in dwarf goats and associated changes in heart rate and heart rate variability

We studied visual discrimination learning in a group of Nigerian dwarf goats using a computer-based learning device which was integrated in the animals home pen. We conducted three consecutive learning tasks (T1, T2 and T3), each of which lasted for 13 days. In each task, a different set of four visual stimuli was presented on a computer screen in a four-choice design. Predefined sequences of stimulus combinations were presented in a pseudorandom order. Animals were rewarded with drinking water when they chose the positive stimulus by pressing a button next to it. Noninvasive measurements of goats heartbeat intervals were carried out on the first and the last 2 days of each learning task. We analysed heart rate (HR) and heart rate variability (HRV) of resting animals to study sustained physiological effects related to general learning challenge rather than acute excitement during an actual learning session. The number of trials to reach the learning criterion was 1000 in T1, when visual stimuli were presented to the goats for the first time, but decreased to 210 in T2 and 240 in T3, respectively. A stable plateau of correct choices between 70% and 80% was reached on Day 10 in T1, on Day 8 in T2 and on Day 6 in T3. We found a significant influence of the task and of the interaction between task and day on learning success. Whereas HR increased throughout T1, this relationship was inverted in T2 and T3, indicating different effects on the HR depending on how familiar goats were with the learning task. We found a significant influence of the task and the interaction between task and time within the task on HRV parameters, indicating changes of vagal activity at the heart. The results suggest that changes in HR related to learning were predominantly caused by a withdrawal of vagal activity at the heart. With regard to nonlinear processes in heartbeat regulation, increased deterministic shares of HRV indicated that the animals did not really relax until the end of T3. Comparing changes of HR and HRV in T3 and in a subsequent postexperiment (PE), we assume a positive effect of such cognitive challenges once the task had been learned by the animals.

Autonomic reactions indicating positive affect during acoustic reward learning in domestic pigs

Cognitive processes, such as stimulus appraisal, are important in generating emotional states and successful coping with cognitive challenges is thought to induce positive emotions. We investigated learning behaviour and autonomic reactions, including heart rate (HR) and its variability (standard deviation (SDNN) and root mean square of successive differences (RMSSD) of a time series of interbeat intervals). Twenty-four domestic pigs, Sus scrofa, housed in six groups of four, were confronted with a cognitive challenge integrated into their familiar housing environment. Pigs were rewarded with food after they mastered the discrimination of an individual acoustic signal followed by an operant task. All pigs quickly learned the tasks, while baseline SDNN and RMSSD increased significantly throughout the experiment. In reaction to the signals, pigs showed a sudden increase in HR, SDNN and RMSSD, and a decrease in the RMSSD/SDNN ratio. Immediately after this reaction, the HR and SDNN decreased, and the RMSSD/SDNN ratio increased. During feeding, the HR and the RMSSD/SDNN ratio stayed elevated. The pigs showed no cardiac reaction to the sound signals for other pigs or their feeding pen mates. We concluded that the level of cognitive challenge was adequate and that the observed changes in the autonomic tone, which are related to different dimensions of the affective response (e.g. arousal and valence), indicated arousal and positive affective appraisal by the pigs. These findings provide valuable insight into the assessment of positive emotions in animals and support the use of an adequate cognitive enrichment to improve animal welfare.

Linear prediction coding analysis and self-organizing feature map as tools to classify stress calls of domestic pigs (Sus scrofa)

It is assumed that calls may give information about the inner (emotional) state of an animal. Hence, in the last years sound analysis has become an increasingly important tool for the interpretation of the behavior, the health condition, and the well-being of animals. A procedure was developed that allows the characterization, classification, and visualization of the cluster structures of stress calls of domestic pigs (Sus scrofa). Based on the acoustic model of the sound production the extraction of features from calls was performed with linear prediction coding (LPC). A vector-based self-organizing neuronal network was trained with the determined LPC coefficients, resulting in a feature map. The cluster structure of the calls was then visualized with a unified matrix and the neurons were labeled for their input origin. The basic applicability of the procedure was tested by using two examples which were of special interest for a possible evaluation of the normal farming practice. The procedure worked well both in discriminating individual piglets by their scream characteristics and in classifying pig stress calls vs other calls and noise occurring under normal farming conditions.

Effects of attention and rewarded activity on immune parameters and wound healing in pigs.

Challenging animals with a demanding situation they are able to cope with and where they are rewarded may be a source of positive emotions and possibly of increased welfare. In order to test if this results in changes of immunological parameters and wound healing, 56 pigs (7-20 weeks of age), housed in groups of 8 animals each, were successfully trained to recognize and localize an individual acoustic summons and to receive a small portion of feed as a reward. Immune reactions and the development of a standardized biopsy wound were compared to values of conventionally fed control groups of equal size and animal-to-feeding-place ratio (2:1). In the experimental animals a significantly higher concentration of IgG as well as an increased in vitro T-cell proliferation to ConA but a reduced LPS-induced proliferation of B-cells was found, while basal salivary cortisol concentrations were similar. Wound development was better in the experimental animals as measured by the area of the inflammatory corona. It decreased more rapidly in the experimental animals since the 5th day after biopsy and was significantly smaller than in the control groups. We conclude that environmental enrichment by equipment provoking attention and cognitive activity which is rewarded by feed may play a beneficial role for physical welfare of intensively housed pigs.

Common features and individual differences in nurse grunting of domestic pigs (Sus scrofa) : A multi-parametric analysis

The nursing vocalisation of domestic pigs was investigated with respect to common features and individual differences. The sows repeated grunting during nursing can be regarded as a contact call and a signal of the mother to start and synchronise the suckling behaviour of the piglets. Analyses in the time domain revealed the gross structure of the call, whereas in the frequency domain the fine structure of single grunts was investigated. Three parameter groups with parameter vectors extracted from single grunts centred around the maximum of the grunting rate were used for classification of frequency patterns. The parameter vectors were compared by a discriminant function analysis. The main results are: (1) a strong correlation in the time course of the nurse grunting between the sows; (2) an individual frequency pattern of the single grunts as demonstrated by the discriminant analysis with erroneous discriminations of less than 1.6% if an optimum set of features of the amplitude spectrum was included. Thus, it can be concluded that a common structure of the time course of the nurse grunting is the typical inter-individual characteristic of the vocalisation. Individual differences between sows occurred in the absolute grunting rate, however. Major individual differences were found in the frequency pattern of the single grunts and were expressed over the whole frequency range of a grunt. Hence, based on the parameter groups whole amplitude spectrum and cepstrum, individual sows could well be distinguished whereas this was not the case using only single features of the amplitude spectra. Taken together, the results show that there is a species-typical pattern in the nurse grunting that could be recognised by all piglets, and an individual label that could be used to discriminate the own mother from other nursing sows.

Behavioural and cardiac responses towards conspecific distress calls in domestic pigs (Sus scrofa).

In domestic pigs, vocalisation can be an indicator of distress and negative emotional states. It might play a role in the transfer of emotion between individuals (emotional contagion or empathy), which could result in impaired animal welfare on a group level based on the distress in an individual member of the group. The aim of this study was to characterise the responses of pigs to conspecific distress calls. We performed a playback experiment in an open arena with 24 juvenile German Landrace pigs, during which individual subjects were exposed to both conspecific distress calls and an artificial sine tone (control) on consecutive experimental days. Both behavioural (locomotion, vocalisation, elimination and distance to the speakers) and physiological responses (heart rate and heart rate variability) were measured for 2 min before, during and after the playback of the stimuli (distress calls/control). Subjects showed decreased locomotion and vocalisation rates during both stimuli, suggesting that the animals responded to both stimuli. Heart rates decreased at the onset of both stimuli due to an activation of the parasympathetic system, indicating an orientation response to sudden stimuli. However, heart rates decreased after the end of the distress calls but not after control stimuli, illustrating that conspecific calls and other sounds are evaluated differently. We conclude that pigs exposed to isolation are attentive to conspecific distress vocalisation and hence the information about threat possibly conveyed in it, but they do not share the distress of the caller. Therefore, we could not find direct effects of distress calls of unfamiliar conspecifics on the welfare of isolated juvenile pigs. However, the state of heightened attention elicited by conspecific distress calls may affect a pigs subsequent evaluation of its environment.

Individual calling to the feeding station can reduce agonistic interactions and lesions in group housed sows.

In this study we used a new call feeding station, which enabled sows to learn that they have access to feed only after an individual acoustic signal was given. We tested whether this call feeding station is able to reduce agonistic interactions between sows and whether effects of call feeding can further be improved by enrichment. A total of 85 gestating sows were kept in a dynamic group in a large waiting area (207 m²) equipped with littered laying areas and an outside area. During a control treatment sows were fed in a normal electronic feeding station once a day (NF1-). Before testing the call feeding station sows had been conditioned for an acoustic signal (a trisyllabic name) and learned that they were allowed to enter the feeding station only after their name was called. In the call feeding station sows were fed either once (CF1-) or twice a day (CF2-). In addition, we tested for effects of further enrichment such as straw in the activity area (CF2+). Agonistic behaviors and number of sows were observed by video in continuous recording from 0600 to 1800 h in an area (4 by 3.25 m) in front of the feeding station in periods of 4 d (NF1- = 7 periods, CF1- = 5 periods, CF2- = 3 periods, and CF2+ = 4 periods) and analyzed using mixed models. During each observation period sows were scored for wounds and body lesions at different body parts. From 0600 to 1100 h the proportion of agonistic interactions was much greater in the feeding mode NF1- compared with CF1- (feeding mode × time of day: P < 0.001) and in CF1- agonistic interactions were on a low level throughout the whole feeding cycle. The feeding frequency and the additional presence of straw in the activity area did not affect the proportion of agonistic interactions (all P > 0.05). The results on the number of sows in front of the feeding station mirrored the findings for agonistic interactions. In NF1- more sows were involved in agonistic interactions compared with CF1- (83.1 ± 12.9% vs. 61.5 ± 19.6%; P = 0.005) but there was no difference between the call feeding station (CF) treatments. The number of severe lesions was greater at the head (P = 0.014) and the flank (P = 0.006) but not at the shoulders (P = 0.057) and the hindquarter (P = 0.426) in NF1- compared with CF1-. The results suggest that signaling the feeding time individually increases the predictability for access to the feeding station and consequently reduces competition between sows.

An automated training device for pattern discrimination learning of group-housed gerbils

The setup, designed for the rodent Meriones unguiculatus (gerbil), allows flexible stimulus presentations and rewarding as well as on-line data registration. It consists of a spacious housing where the animals have free access to water. Food is supplied exclusively in the y-maze training compartment and serves as a reinforcer in an operant conditioning paradigm.