Identification of Behaviour of Weaner Pigs in the Early

Phase of Aggressive Interaction for the Development of

an Automatic Monitoring System of Pig Aggression

Maciej Oczak

, Gunel Ismailova

Lilia Thays Sonoda

, Michaela Fels

Joerg Hartung

, Marcella Guarino

, Stefano Viazzi

, Daniel Berckmans

and

Erik Vranken

Fancom Research, Panningen, The Netherlands

Institute for Animal Hygiene, Animal Welfare and Farm Animal Behaviour,

University of Veterinary Medicine Hannover Foundation, Hannover, Germany

Department of Veterinary Science and Technologies for Food Safety

Faculty of Veterinary Science, Milan, Italy

M3-BIORES: Measure, Model & Manage Bioresponses,

Katholieke Universiteit Leuven, Leuven, Belgium

Abstract. Aggressive behaviour among pigs results in negative consequences,

reducing health and welfare of animals as well as production output of a farm.

In order to develop an automatic monitoring system that monitors and controls

pig aggression, an experiment was carried out. The aim of the experiment was

to analyze sequences in pig aggressive behaviour. 52 aggressive sequences were

observed during the experiment in on farm observations. Behaviour that started

the aggressive sequences the most often was nose to nose interaction. 22 out of

52 aggressive sequences started with this behaviour. Head to head knocking

was classified as a second most frequent starting behaviour with 13 head to

head knocks starting aggressive interactions. Nose to nose interaction and head

to head knocking behaviours in most cases started aggressive interactions be-

tween animals. Automatic detection of these behaviours might allow early de-

tection of aggression among pigs.

1 Introduction

The social organization of domesticated pigs, Sus scrofa, living under farm conditions

is, as in wild pigs, based upon a dominance hierarchy [10]. Hierarchy is established

when unacquainted pigs are placed together [13], [4]. This common commercial prac-

tice results in intense aggression which extends throughout the first 24-48h after

grouping [9]. The reasons for increased aggression level among pigs in confinement

are: limited space allowance [15], feeding systems promoting competition [9], barren

environment [2], low fibre feed composition [12] and regrouping [9].

The fight breaks out gradually as the pigs investigate each other using a series of

specific and often reciprocal behaviours, characterized by nosing, sniffing and gentle

nudging. This may then escalate into more vigorous pushing and pressing, bites, head-

knocking and mounting. Thus, as the fight progresses, it increases in intensity (i.e.,

more damaging behaviours occur more frequently later in the fight). A complete un-

Oczak M., Ismailova G., Thays Sonoda L., Fels M., Hartung J., Guarino M., Viazzi S., Berckmans D. and Vranken E..

Identiﬁcation of Behaviour of Weaner Pigs in the Early Phase of Aggressive Interaction for the Development of an Automatic Monitoring System of Pig

Aggression.

DOI: 10.5220/0003885700240029

In Proceedings of the International Workshop on Veterinary Biosignals and Biodevices (VBB-2012), pages 24-29

ISBN: 978-989-8425-94-2

 2012 SCITEPRESS (Science and Technology Publications, Lda.)

derstanding of aggressive behaviour may help in improvement of livestock manage-

ment, as well as health and welfare of animals [11].

Aggressive behaviour has a negative impact on health, welfare [10], and produc-

tivity of pigs [1]. At present evaluation of health, welfare and productivity is per-

formed by stockmen who routinely gather auditory, olfactory and visual information

from their animals. New precision livestock farming technology can aid this task,

even with large flocks or herds, thanks to the (r)evolution in sensors and sensing tech-

niques [14]. Therefore there is an opportunity for application of precision livestock

farming technology as an automatic aggression monitoring system.

The objective of this paper is to present the development of automatic monitoring

system of pig aggression, experimental installation utilized in the development proc-

ess and analysis of pig’s aggressive behaviour.

2 Application of PLF as an Automatic Monitoring System of Pig

Aggression

Precision livestock farming (PLF), can be defined as the management of livestock

production using the principles and technology of process engineering. It is the prin-

cipal means by which ‘smart’ sensors will be used in livestock farming [14]. An inte-

grated monitoring system is one which collects information from a variety of sources,

including sensors, databases and knowledge bases, processes the data and provides

outputs, which may be recommendations to the producer, or direct process control

actions [5].

Fig.1 presents the PLF scheme adapted to automatic monitoring system of pig ag-

gression. The scheme is being realized by model predictive control. This does not

prescribe a specific control strategy, but rather a range of control methods, which use

continuous feedback of the process output (as in other control strategies), make an

explicit use of a dynamic model of the process to predict the process response, and

use this model to calculate the control signal by minimizing an objective function

[14].

The first step in aggression monitoring system development is the development of

an automatic monitor which by dynamic analysis of the inter-individual interactions

between animals, will be able to detect aggressive attacks automatically and identify

the aggressor(s) and the receiver(s) in a group. The second step is to develop a control

strategy. Control strategy has to be realized by utilization of actuator (i.e. sound,

smell) that changes pig behaviour, lowering aggression level.

In order to develop monitoring system and control strategy for pig’s aggressive

behaviour in depth analyzes of these behaviours is necessary. To obtain data neces-

sary in product development process an experimental installation was constructed and

an experiment carried out.

Fig. 1. Schematic overview of key components of automatic monitoring system.

3 Materials and Methods

3.1 Ethological Observation Installation

Behaviour of one pen of 11 pigs was observed on farm by experts in pig’s behaviour

and recorded with video camera (Fig.2). Recordings were done using an Allied Vision

Technologies® camera model F080C. The camera was placed above the observed pen

in central position (in top view), at the height of 2.3m, connected to a computer which

recorded the images with a frame rate of 11 fps, resolution of 1032 x 778 pixels and

in colour.

The video recordings and visual observations were carried out in the same time, in

the period of 3 days after the pigs were mixed. On the first day of the observation 35

minutes (16:50 – 17:25) of videos were recorded. On the second day 3 hours and 30

minutes (13:00 – 16:30) and on the third day 3 hours (11:00 – 14:00) were recorded.

Behavioural recordings were carried out on 11 pigs of 23 kg on average kept in a pen

of 4m x 2,5m with partially slatted concrete floor and solid pen walls. The feeder bin

in the pen had 2 feeding places.

In parallel with video recordings there were on-farm visual observations per-

formed. The requirements for visual observation were reached in an installation con-

sisting of a sheet of strong paper hanging in the front of the pen and a platform placed

behind the paper wall. The observers were sitting on a platform (around 150 cm above

ground level) in a comfortable position, high enough to be able to see the whole sur-

face of the pen. At the height of the observers’ eyesight there was an opening cut in

the sheet of paper (40x30 cm). The size of the opening was calculated to be big

enough for the observer to see the whole area of the pen, but simultaneously, rela-

tively small so pigs couldn’t see too much movement behind the paper. The usage of

such a paper wall was motivated by the fact that too much movement around the pen

could cause distraction of the pigs and changes in their behaviour. During the real

time labelling, behaviours were presented in a database with the time and the se-

quence of events that occurred inside the pen.

On farm observations data and videos recorded during the experiment were used

to specify agonistic behavioural patterns occurring among pigs.

Fig. 2. Labelling process.

3.2 Pigs Behaviour

Data obtained by experts during on farm observations were used to specify the order

in which aggressive behaviours occur. Labelling of recorded videos gave more de-

tailed information and allowed to analyze the pigs body postures, relation of body

positions, activity and occupation indexes. Comparison of the on farm observations

and the video recordings was also performed (Fig.2).

Behaviours observed during on farm observations were recorded in an ethogram.

The ethogram contained following behaviours [6], [3]:

Nose to nose: The nose approaches the snout or head of the receiver.

Head to head knock: Hitting with the snout against the head of the receiver.

Inverse parallel pressing: Pressing of shoulders against each other, facing opposite

directions.

Neck biting: Biting of the opponent pig directed towards the neck.

Body biting: Biting of the opponent pig directed towards the body.

Mounting: Jumping on another pig with legs directed forward.

Ear manipulation: Holding receiving pig’s ear in the mouth.

4 Results and Discussion

4.1 On Farm Observations

52 aggressive sequences were observed during 3 days of the experiment. Behaviour

that started the aggressive sequences the most often was nose to nose interaction. 22

out of 52 aggressive sequences started with this behaviour. Head knocking was classi-

fied as a second most frequent starting behaviour with 13 head knocks starting ag-

gressive interactions. Neck biting started 7 out of 52 aggressive interactions.

The behaviours that started the aggressive sequences the least often were: inverse

parallel pressing (4 out of 52), mounting (2 out of 52), body biting (2 out of 52), and

ear manipulation (2 out of 52) (Fig. 3).

Fig. 3. Behaviours that started the aggression.

4.2 Discussion

In our experiment sequences of aggressive behaviour mostly started with nose to nose

interaction, which is similar to Jensen’s findings [7-8].

The second most frequently occurring as starting behaviour was head to head

knocking behaviour. The behaviour classified by Jensen [8] as one of the behaviours

always occurring after nose to nose interaction, in our experiment was often (13 of

52) an aggression initiating behaviour.

Neck and body biting behaviours occurred less often at the beginning of aggres-

sive sequences which complies with findings of other scientists [7-8], [11]. McGlone

[11] showed in his study that mutual bites occur during the late phase of the fight;

therefore these behaviours usually don’t initiate aggressive sequences.

5 Conclusions

In the development of an automatic monitoring system of pig aggression it is neces-

sary to identify behaviours that can be recognized by the system and monitored con-

tinuously. Nose to nose interaction and head to head knocking are behaviours identi-

fied as occurring at the beginning of aggressive sequences. Therefore automatic de-

tection of these behaviours might allow early detection of aggression between pigs.

Aggressive

sequences

22 Nose

to nose

13 Head

to head

knocks

7 Neck

biting

2 Body

biting

Mounting

4 Inverse

parallel

pressing

2 Ear

manipu

lation

References

1. Arey D. S., Edwards S. A., (1998). Factors influencing aggression between sows after

mixing and the consequences for welfare and production. Livestock Production Science 56:

61–70.

2. Durrell, J. L., Sneddon, I. A., Beattie, V. E. (1997). Effects of enrichment and floor type on

behaviour of cubicle housed sows. Anim Welfare 6: 297–308.

3. Fraser, D., (1974). The behaviour of growing pigs during experimental social encounters. J.

Agric. Sci., 82: 147-163.

4. Friend T. H., Knabe D. A. and Tanksley T. D., (1983). Behaviour and performance of pigs

grouped by three different methods at weaning. J. Anim. Sci., 57: 1406–1411.

5. Frost AR., Schofield C. P., Beaulah S.A., Mottram T. T., Lines J. A., Wathes C.M., (1997)

A review of livestock monitoring and the need for integrated systems. Computers and Elec-

tronics in Agriculture 17: 139-159

6. Gonyou, H., (2001). The social behaviour of pigs. In: Keeling, L. and H. Gonyou (editors),

Social behaviours in farm animals.

7. Jensen P., Yngvesson J., (1998). Aggression between unacquainted pigs—sequential as-

sessment and effects of familiarity and weight. Applied Animal Behaviour Science 58: 49–

61.

8. Jensen, P., (1982). An analysis of agonistic interaction patterns in group-housed dry sows -

aggression regulation through an “avoidance order”. Appl. Anim. Ethel., 9: 47-61.

9. Marchant-Forde J. N, (2010). Social behaviour in swine and its impact on welfare. Proceed-

ings of the 21st IPVS Congress, Vancouver, Canada 18-21

10. McBride G. and James J.W., (1964). Animal production. Social behaviour of domestic

animals. IV. Growing pigs. Animal Production, 6: 129–139.

11. McGlone, J. J. (1985) A quantitative ethogram of aggressive and submissive behaviours in

recently regrouped pigs. J Anim Sci 61: 559-565.

12. Meunier-Salaun, M. C., Edwards, S. A., Robert, S. (2001). Effect of dietary fibre on the

behaviour and health of the restricted fed sow. Anim Feed Sci Tech 90: 53–69.

13. Sherritt G.W., Graves H.B., Gobble J.L. and Hazlett V.E., (1974), Effects of mixing pigs

during the growing–finishing period. J. Anim. Sci., 39: 834–837.

14. Wathes C.M. , Kristensen H.H. , Aerts J.-M. , Berckmans D., (2008). Is precision livestock

farming an engineer's daydream or nightmare, an animal's friend or foe, and a farmer's pan-

acea or pitfall? Computers and Electronics in Agriculture. Volume 64: 2-10. Smart Sensors

in precision livestock farming.

15. Weng R. C., Edwards S. A., English P. R., (1998). Behaviour, social interactions and lesion

scores of group-housed sows in relation to floor space allowance. Applied Animal Behav-

iour Science. 59: 307-316.