ELSEVIER Applied Animal Behaviour Science 53 (1997) 33-43

APPLIED ANIMAL BEHAVIOUR

SCIENCE

The contribution of applied ethology in judging animal welfare in farm animal housing systems

Beat Wechsler a3 * , Ernst Friihlich b, Hans Oester b, Thomas Oswald ‘, Josef Troxler ‘, Roland Weber d, Hans Schmid e a Abteilung Sozial- und Nutztierethologie, Zooiogisches Institut, Uniuersitiit Bern, Ethologische Station Hash,

Wohlenstrasse 5Oa, 3032 Hinterkappelen, Switzerland b Bundesamt fur Veterindnvesen, Priifstelle fir Stalleinrichtungen, Burgerweg 22, 3052 Zollikofen,

Switzerland ’ Bundesamt jiir Veteriniirwesen, Priifstelle fir Stalleinrichtungen, Eidg. Forschungsanstaltfir

Agranvirtschaft und Landtechnik 83.56 Tdnikon, Switzerland ’ Eidgeniissische Forschungsanstalt fur Agrarwirtschaft und Landtechnik, Priifstelle fur Stalleinrichtungen,

8356 Tiinikon, Switzerland ’ Tiergartenbiologische Abteilung, Zoologisches Institut, Uniuersitiit Zurich, Ziirichbergstrasse 221, 8044

Zurich, Switzerland

Abstract

In 1981, the Swiss animal welfare legislation introduced an authorisation procedure for mass-produced farm animal housing systems. If appropriate, practical tests are carried out to identify animal welfare problems in systems or devices that are proposed by manufacturers. Such tests may include veterinary, physiological and behavioural measurements. We briefly review practical tests of group cages for laying hens, electric cow-trainers and farrowing crates for sows. These examples illustrate how ethological studies make important contributions in judging animal welfare. It is also stressed that the development of alternative housing systems that are adapted to the behavioural organisation of farm animals should be a major field of applied ethology. 0 1997 Elsevier Science B.V.

Keywords: Legislation: Laying hen group cages; Electric cow-trainers; Farrowing crates

1. Introduction

The scientific assessment of animal welfare is not a simple task (recent reviews by Rushen and de PassillC, 1992; Mason and Mend], 1993; Fraser, 1995). First, there is a

* Corresponding author. Swiss Federal Research Station for Agricultural Economics and Engineering, 8356 rinikon, Switzerland, Tel.: 052 368 31 31; fax: 052 365 11 90; e-mail: beat.wechsler@fat.admin.ch

0168-1591/97/$17.00 0 1997 Elsevier Science B.V. All rights reserved. PII SOl68-1591(96)01149-5

34 B. Wechsler et al. / Applied Animal Behaaiour Science 53 (1997) 33-43

philosophical problem in that no one, not even a scientist, has access to the private experiences of animals (McFarland, 1989, pp. 34-58; Bateson, 1991). In practice, animal welfare is therefore assessed by means of indicators that are assumed to be related to the subjective feelings of animals (Oester, 1982; Troxler and Steiger, 1982; Sandoe and Simonsen, 1992). Common indicators are measurements of mortality, disease, injury, productivity, physiology, immunology and behaviour (Dawkins, 1980; Broom, 1991). Second, however, there may be problems with the validity and interpreta- tion of these measures (Moberg, 1987; Barnett and Hemsworth, 1990; Rushen, 1991; Mason and Mendl, 1993). Third, it is difficult to justify the definition of a cut-off point for determining when an animal’s welfare is good or poor (Duncan and Dawkins, 1983; Mendl, 1991). Decisions about the acceptance or banning of specific housing systems (e.g. battery cages for laying hens, housing of calves in individual crates, farrowing crates for sows) will therefore not only be based on the scientific assessment of welfare but also on ethical and political considerations (Tannenbaum, 1991; Rushen and de PassillC, 1992; Sandae and Simonsen, 1992; Fraser, 1995).

Despite these philosophical and scientific considerations it is generally accepted by society that it is possible to make judgements about ‘well-being’ and ‘suffering’ in animals. Consequently, animal welfare legislation includes requests to safeguard the ‘well-being’ of animals or to prevent them from ‘suffering’, and scientists may be confronted with the task of evaluating the welfare of animals kept in a given housing system.

The aim of our review is to illustrate how such tasks are dealt with in Switzerland. We will first describe the legal and theoretical background for the practical examination of housing systems in our country. We will then focus on studies that have been done to evaluate group cages for laying hens, electric cow-trainers and farrowing crates. Thereby, we will emphasise the significance of behavioural measures in identifying welfare problems in farm animals.

2. Swiss animal welfare legislation

In 1978, the Swiss people voted for new animal welfare legislation (the Animal Welfare Act), and in 1981 the Federal Council put into force detailed regulations (the Animal Protection Regulations). The principles of the Swiss Animal Welfare Act are listed in article 2. It is stated that: 1. “animals shall be treated in the manner which best accords with their needs”; 2. “anyone who is concerned with animals shall, in so far as circumstances permit,

safeguard their welfare”; 3. “no one shall unjustifiably expose animals to pain, suffering, physical injury or

fear”. Obviously, those who wrote this legislation were convinced that it is feasible to assess ‘needs’, ‘suffering’ and ‘fear’ in animals. Article 1 of the Swiss Animal Protection Regulations contains the following general

statements about housing conditions: 1. “animals shall be kept in such a way as not to interfere with their bodily functions or

their behaviour, or overtax their capacity to adapt”;

B. Wechsler et al. /Applied Animal Behaciour Science 53 f 1997133-43 35

2. “feeding, care and housing shall be deemed suitable where, according to existing experience and the state of knowledge concerning physiology, animal behaviour and hygiene, they comply with the animals’ requirements”.

Both statements mention animal behaviour as an important factor for proper housing. As a consequence, applied ethology is expected to be able to judge when an animal’s capacity to adapt to a housing system is overtaxed.

In line with this expectation, Article 5 of the Swiss Animal Welfare Act prescribes an authorisation procedure for the sale of housing systems for farm animals (i.e. cattle, sheep, goats, pigs, rabbits and poultry). It is required that:

mass-produced housing systems and installations for the keeping of animals for purposes of profit may not be advertised and sold without prior authorisation from a service designated by the Federal Council. Authorisation shall only be granted if such systems and installations provide proper living conditions for animals.

Housing systems and installations that were already on the market in 1981 may continue to be sold unless they are prohibited. Authorisations are given by the Federal Veterinary Office. It may consult an advisory board which consists of three representa- tives of the Confederation, three representatives from the cantons, four scientists, three experts in animal protection, three experts in the keeping of animals and three experts in animal housing construction. The Federal Veterinary Office can put a time limit on the validity of an authorisation and impose specific requirements concerning the use of a housing system or installation. The manufacturers may appeal against a decision of the Federal Veterinary Office.

3. Testing of farm animal housing systems

The authorisation procedure applies not only to complete housing systems such as cages, boxes or crates but also to installations with which animals frequently come in contact (e.g. feeding and watering systems, floor coverings, dung grids, tethering arrangements, nest boxes). When asking for an authorisation, the manufacturer or importer of a housing system or installation must send detailed documentation (plans, measures, technical data) to the Federal Veterinary Office. The authorisation can only be given if the housing system or installation is in accordance with the requirements of the Swiss animal welfare legislation. Whenever possible, a decision is made on the basis of literature or experience with similar installations. In some cases, however, practical tests are required. Two research stations have been established by the Federal Veterinary Office for this purpose. Housing systems for cattle, goats, sheep and pigs are tested in T’anikon, those for domestic poultry and rabbits in Zollikofen. Sometimes such tests are also carried out on commercial farms where the housing systems or installations under analysis are in use under temporary authorisation.

A practical test may include veterinary, physiological and ethological indicators of animal welfare. It is carefully designed with respect to the sample size and the distribution of data collection over time. The latter is done to identify both short- and long-term welfare problems of a given housing system or installation. If appropriate, the

36 B. Wechsler et al. /Applied Animal Behauiour Science 53 (1997) 33-43

tested housing system is compared with a reference system (e.g. farrowing crate versus farrowing pen). For the authorisation procedure the results of a practical test are supplemented with evidence found in the literature.

The theoretical background of the practical examination of housing systems in Switzerland was much influenced by Beat Tschanz, professor in ethology at the University of Berne until 1987. Considering the fact that there is no direct access to the ‘well-being’ or ‘suffering’ of animals he focused on characteristics that can reliably be measured by different observers (Tschanz, 1982, 1987). He argued that all animals aim at satisfying their requirements (e.g. food, water, shade) and at avoiding harm in order to enhance survival and reproduction. If an animal cannot achieve these two functions in a housing system, its capacity to adapt is obviously overtaxed, and relevant deficiencies (e.g. malnutrition, injuries) can be measured objectively. Tschanz also emphasised the significance of behaviour in achieving these functions. Well before pathology occurs, modifications in behaviour may indicate that an animal has difficulties in adapting to its environment. A practical test of a housing system should therefore not only include veterinary and physiological measures but also behavioural observations. In the follow- ing, we report some ethological studies that have been carried out during authorisation procedures. In each example, we briefly review the methods, the main results and the decisions made by the authorities.

4. Group cages for laying hens

According to the Swiss Animal Protection Regulations, housing systems for laying hens must contain protected, darkened and soft-floored or litter-lined nest boxes, as well as suitable perches or a slatted floor. For housing systems with a gridded floor the prescribed minimum floor area per hen is 1400 cm2 for groups of up to ten hens, 1200 cm2 for groups of 1 l-20 hens, 1000 cm2 for groups of 21-40 hens and 800 cm2 for groups of more than 40 hens. As conventional battery cages do not fulfil these requirements, they had to be replaced by alternative housing systems within a transi- tional period of 10 years which ended in 1991.

The regulations for laying hens do not, however, explicitly ban cage housing systems. Consequently, several manufacturers asked for authorisation to mass-produce and sell modified cages for groups of 40 or more hens. Two of these group cages, one with 42 hens per cage (Oekonom 165) and one with 58 hens per cage (Scharbo 162 V), were tested on commercial farms and at the research station in Zollikofen. The farms were visited when the hens were 18, 30 and 50 weeks old, and at the end of the egg-laying period. Data collected included scores of plumage condition, measurements of the length of the claws, recordings of injuries and mortality as well as observations on the behaviour of the hens. A total of 101 (Oekonom 165) and 132 (Scharbo 162 V) cages installed on three and five farms, respectively, were involved in the testing. Detailed observations on resting, aggressive behaviour, feather pecking and cannibalism were made in 252 and 176 hens housed in six (Oekonom 165) and three (Scharbo 162 V> cages, respectively, at the research station. Both types of group cages were arranged in three tiers with automatic removal of droppings. They contained nest boxes with a

B. Wechsler et al. /Applied Animal Behauiour Science 53 (1997) 33-43 37

sloped plastic floor. The feeding troughs were in front of the cages and the drinkers in the middle. The cages for groups of 42 hens had a slatted plastic floor whereas the cages for groups of 58 hens had a wire-mesh floor with perches. Neither system contained litter.

The behavioural observations revealed several welfare problems. First, the hens seemed not to differentiate between resting and activity areas. Resting bouts of individ- ual hens were regularly interrupted by other hens that moved to the feeding trough or the nipple drinkers. Second, the hens frequently paced along the boundaries of the cages during the daily egg-laying period, indicating difficulties with the selection of a suitable nest-site. Third, the hens had no adequate substrate to perform dustbathing behaviour, as there was no litter in the cages. Vacuum dustbathing was regularly observed on the floor, and bill-raking (which hens usually show at the start of a dustbathing bout) was often directed at the food in the trough. Fourth, the absence of elevated perches prevented the hens from finding an appropriate place to rest at night. There was still much activity in the cages when the lights had been turned off. Fifth, normal foraging behaviour with scratching and ground-pecking was impaired because of the lack of litter. The hens showed bouts of repetitive pecking at parts of the cage as well as bouts of alternating pecking at food and at parts of the cage. Sixth, there were serious problems with feather-pecking. Although the average light intensity within the cages was less than 5 lx, the plumage condition deteriorated much with age. Seventh, over a period of 56 weeks, 10.9% (Scharbo 162 V) and 16.7% (Oekonom 165) of the hens kept in group cages at the research station died due to cannibalism.

Based on the results of the practical tests, group cages for 40-60 laying hens were judged to overtax the hens’ capacity to adapt (Frohlich and Oester, 1989). The Federal Veterinary Office did not therefore permit the manufacturers to mass-produce and sell such modified cages. On the other hand, in 1986 authorisation was granted for the first five aviary systems that had been subjected to practical tests (Oester and Frijhlich, 1986). Today, 14 different aviary systems are on the market.

5. Electric cow-trainers

In Switzerland, only a minority of cows are kept in loose housing systems. Most cows are still tethered in a traditional way, and it is usual to combine this type of housing system with the cow-trainer. This consists of an electrified metal rod fixed in a horizontal position a few centimetres above the withers of the tethered cow. Cattle arch their backs when urinating or defecating. This arching will not result in their touching the cow-trainer if they move back sufficiently, so that the rod is above their neck rather than their withers. As a consequence, the cow’s lying place is kept clean with little labour.

The use of the cow-trainer is not prohibited by the Swiss Animal Protection Regulations, but this did not exempt manufacturers of cow-trainers from having their products tested at the research station in Tanikon (Oswald, 1992). In a detailed study, the behaviour of dairy cows was observed with and without the cow-trainer. Eight cows that had experience with a cow-trainer for at least one year were kept without a

38 B. Wechsler et al. /Applied Animal Behauiour Science 53 (1997) 33-43

Table 1 Mean (+ SD) frequencies (per day) of behavioural elements by which the cows (N = 8) either broke a beam of light or touched a cow-trainer that was 5 cm above their withers

Behaviour Beam of light Cow-trainer Pa

Eat Stand Fly removal Lick back Other comfort behaviour Stand up Defecate Urinate

13.9+ 13.2 5.9+3.8 3.6+2.1

12.5+5.8 6.5k3.2 4.8+2.9 3.2k2.5 4.0+ 2.4

1.0+ 1.1 1.4kO.7 0.7 +0.7 0.9* 1.0 1.8+ 1.8 0.9kO.7 0.2 + 0.3 0.6 + 0.6

0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05

a Wilcoxon matched-pairs signed-ranks test.

cow-trainer for 8 weeks. They were then moved to an experimental stall where a beam of light trained on a photocell was placed 5 cm above the withers. After a week of acclimatisation, all behavioural elements by which the cow broke the beam of light were recorded continuously for 120 h. Subsequently, a cow-trainer was fixed parallel to the beam of light, and after another week of acclimatisation, all behavioural elements that led to a contact with the cow-trainer were recorded for 120 h.

The cow-trainer had a marked effect on various behavioural elements (Table 1). The unrestricted cows frequently broke the beam of light, especially while eating or licking the back. When the cow-trainer was in use, the cows were obviously careful not to raise their back or their head (to lick the back or remove flies) too high. It was also found that only 11% of all contacts with the cow-trainer happened during urinating or defecating. Based on the reactions of cows that received electric shocks, Oswald (1992) judged that the cow-trainer may cause pain, suffering and states of fear. In his final report on the practical tests he concluded that from an ethological point of view the cow-trainer is not compatible with basic requirements of the Swiss animal welfare legislation.

In spite of this conclusion, the Federal Veterinary Office gave the manufacturers permission to sell cow-trainers. It was argued that there are no practicable alternatives to keep the tether stalls and the cows clean. The only alternative would consist in having much shorter stalls, but these are known to result in damage to the hind legs and the udders of the cows. The cow-trainer was thus considered as a ‘necessary evil’ in housing systems with tethered cows. In order to minimise the negative effects of the cow-trainer, several requirements were imposed concerning its use. For example, it was prescribed that: 1. the cow-trainer must not be used for stalls shorter than 175 cm; 2. the distance between withers and cow-trainer must be at least 5 cm; 3. the cow-trainer must not be used for animals other than cows or heifers of at least 18

months of age; 4. the cow-trainer must not be on permanently. Oswald (1992) showed experimentally that the cows were not any dirtier if the cow-trainer was electrified for 2 days per week instead of 7.

B. Wechsler et al. /Applied Animal Behar,iour Science 53 (1997133-43 39

6. Farrowing crates

When the Swiss Animal Protection Regulations were enacted in 1981, farrowing crates were already widespread in intensive pig production. Nevertheless, manufacturers who wanted to sell farrowing crates had to ask for authorisation. The practical tests of farrowing crates were conducted from 1984 to 1986 at the research station in Tlnikon. The behaviour of sows kept in farrowing crates was compared with that of loose-housed sows.

Four types of farrowing crates from different manufacturers were tested at the same time. They differed in the size of the pen in which the crate was installed, the location of the piglet heater and the construction of the crate (Weber, 1987). The loose-housed sows were kept in a straw-bedded farrowing pen (2.9 m X 2.4 m). The practical tests included 39 litters in the farrowing pen and 77 litters in the farrowing crates. The sows were introduced into the housing systems about 1 week before the expected farrowing date. The piglets were artificially weaned when they were about 3.5 days old. With six litters in the farrowing pen and 21 litters in the farrowing crates, detailed behavioural observations were made of the sows and the piglets.

Several welfare problems were evident in all four types of farrowing crates. Although the sows were offered straw in the crates (a requirement of the Swiss Animal Protection Regulations) their nest-building behaviour was impaired (Table 2). During the last 6 h before parturition the sows in the farrowing crates showed significantly less straw carrying and pawing, but more manipulative behaviour directed at parts of the housing system (biting, chewing, nosing), than the loose-housed sows. Weber and Troxler (1988) also found that the total duration of the parturition was longer in the farrowing crates (average 237.9 IL- 78.4 min, N = 20) than in the farrowing pen (average 170.1 + 37.8 min, N = 6). During parturition the farrowing crates prevented the sows from turning around and sniffing at new-born piglets, a behaviour that was regularly observed in loose-housed sows. The crates also affected other aspects of the sows’ normal behaviour (Troxler and Weber, 1989). For example, the sows could not leave the nest-site for defecation, they were very much restricted in their locomotory and exploratory be- haviour, and they often struck parts of their bodies against the crates when lying down or standing up.

Based on these results, farrowing crates were judged not to be in accordance with the requirements of the Swiss animal welfare legislation, and the manufacturers were not

Table 2 Mean ( + SD) frequencies of behavioural elements shown by sows kept in farrowing crates (N = 2 I) or in a farrowing pen (N = 6) during the last 6 h before parturition

Behaviour Crate Pen P ‘1

Straw carrying 3.Ok4.9 21.0*24.1 0.05 Pawing 17.3k21.3 63.7 + 46.6 0.05 Rooting 64.6 f 34.2 91.7k62.6 NS Manipulating parts of housing system 41.3+32.5 25.3 + 39.0 0.05

a Mann-Whitney U test.

40 B. Wechsler et al. /Applied Animal Behaviour Science 53 (1997) 33-43

N P

NEST AREA covered with straw piglet box with installed heater and feeder

ACTIVITY AREA feeding trough for the sow straw- hayrack water dispenser rail with gate threshold 10 cm high

- 270cm -

Fig. 1. Farrowing pen designed by Hans Schmid.

given definitive permission to sell their farrowing crates. It was, however, not feasible to ban all farrowing crates, as there were no practicable alternative fan-owing systems at that time. As a consequence, research on new farrowing systems was intensified.

Within a few years a new farrowing pen (Fig. 1) was developed by Hans Schmid from the Zoology Department at the University of Zurich in collaboration with the research station in Xnikon (Schmid and Weber, 1992; Schmid, 1993). The pen is divided into an activity area and a nest area. The latter offers the sow a sheltered nest-site, as it is surrounded by opaque walls and separated from the activity area by the piglet box. The floor of the nest area is covered with straw, and more straw for nest-building is offered in a rack. The design of the pen is adapted to the peri-parturient behaviour shown by free-ranging domestic pigs (Stolba and Wood-Gush, 1984; Jensen, 1989). The sow can choose a nest-site and perform normal nest-building behaviour. She is free to sniff at new-born piglets, and she can leave her nest-site for defecation. In fact, all 53 sows observed during the practical test of the new farrowing pen chose the nest area for parturition, and Schmid (1992) could show that the sows clearly avoided defecation and urination in the nest-site. As part of the practical test, the reproductive performance of sows that were kept in the new farrowing pen was compared with that of sows of the same herd that were kept in conventional far-rowing crates (Table 3). There were no statistically significant differences (P > 0.05) between the two housing systems. In 1992, the authorities decided that the new farrowing pen could be mass-produced and sold.

Table 3 Reproductive performance (mean+SD) of sows kept in a farrowing crate (N = 52 litters) or in the ethologically designed farrowing pen (N = 53 litters)

Behaviour Crate Pen Pa

No. of piglets born alive No. of piglets weaned Mortality until day 32 (% of piglets born alive) piglet weight at day 28 (kg)

a Mann-Whitney U test.

11.6k2.7 11.2f2.4 NS 10.1 f 1.9 9.8 f 2.0 NS 12.2+ 10.1 11.3 f 10.6 NS

8.0f 1.0 7.6* 1.1 NS

B. Wechsler et al./Applied Animal Behmiour Science 53 (1997) 33-43 41

7. Discussion and conclusions

The assessment of animal welfare is a problem that involves both scientific and philosophical elements (Sandoe and Simonsen, 1992). Given the fact that no one has access to the feelings of animals, one could argue that scientists are not able to measure animal welfare and that terms like ‘well-being’, ‘suffering’ or ‘fear’ should be avoided. The experiences with the practical examination of housing systems in Switzerland show, however, that such an argument is based on a false assumption, In practice, politicians, farmers and consumers do not expect scientists to measure animal welfare as if it were an attribute like blood pressure that can be measured directly. They are satisfied if ethologists and physiologists are able to draw conclusions on animal welfare, based on indicators that can be measured in a scientific way. As emphasised by Fraser (1995) animal welfare is a concept that involves values. In accordance with this view, the scientists’ reports on the practical tests of housing systems and installations in Switzer- land not only include data of various measures that are assumed to be related to the feelings of animals but also an interpretation of these data with regard to animal welfare. Both the data and their interpretation help the Federal Veterinary Office to reach decisions on practical actions that should be taken to safeguard the welfare of animals.

The two research stations of the Federal Veterinary Office at which practical test are carried out do not have the resources to do extensive studies with every housing system or installation proposed by manufacturers. As a consequence, the number of animals used for detailed observations within the facilities of these research stations may be rather small. If possible these data are supplemented with observations made on farms where the housing systems or installations under analysis are in use under temporary authorisation. Due to the limited resources available, the testing is also not normally replicated with different breeds or with different configurations (e.g. cow-trainer fixed 5, 7 and 9 cm above the withers). It may, however, become necessary to repeat a practical test if new breeds with markedly different behaviour are introduced into animal production, or if a housing system or installation that was tested is subjected to a significant modification.

In the above examples behavioural data were both relevant and important in judging animal welfare, and we suggest that applied ethology should form a part of the process that leads to legal decisions about animal welfare. The procedure chosen in Switzerland, with practical tests of housing systems that are intended to be mass-produced, proved to be appropriate. With various housing systems it was not possible to write an expert opinion on animal welfare problems that was solely based on evidence from the literature. Specific ethological studies were necessary to decide whether the require- ments of the Swiss animal welfare legislation were fulfilled or not. As a consequence, the authorisation procedure led to an intensification of applied animal behaviour research.

Since there are major differences in the behavioural organisation of different species, it is not possible to use a simple list of indicators to judge animal welfare in all types of housing systems. In the authorisation procedure, the Federal Veterinary Office often based its decisions on the theoretical concept proposed by Tschanz (1982, 1987). It is convincing to state that an animal’s capacity to adapt is overtaxed when it cannot satisfy

42 B. Wechsler et al. /Applied Animal Behaviour Science 53 (1997) 33-43

its basic requirements or avoid harm. This functional approach, however, is unsuitable when abnormal behaviours that are not paralleled by physical deficiencies or injuries (e.g. weaving in horses, pacing in caged laying hens) have to be judged with regard to animal welfare (Wechsler, 1993). In such cases it is necessary to investigate the causal mechanisms that are responsible for the development of the abnormal behaviour (Dawkins, 1983; Jensen and Toates, 1993).

To improve the quality of animal husbandry it is not sufficient to identify welfare problems in a given housing system. In the examples reported here, the legal decisions of the authorities were much influenced by the availability of alternatives to the housing systems being tested. Group cages for laying hens were not allowed to be mass-pro- duced and sold because the practical tests of aviary systems provided better results with regard to animal welfare. With the farrowing crates the Federal Veterinary Office did not give the manufacturers a definitive permission but supported the development of alternative farrowing systems. On the other hand, cow-trainers were permitted although the practical tests clearly indicated welfare problems. In this case, the best long-term strategy is to replace the tethering of cows by loose housing systems. It is clear from these examples that the development of practicable alternative housing systems should be a major field of applied ethology. In recent years, many studies have been published on the proximate causation, motivation and development of behaviour in poultry, cattle and pigs. This knowledge of the behavioural organisation of farm animals must be incorporated into the design of housing systems. In line with Fraser (199.5), we are convinced that applied ethology can make important contributions to identify and solve or prevent animal welfare problems.

References

Bamett, J.L. and Hemsworth, P.H., 1990. The validity of physiological and behavioural measures of animal welfare. Appl. Anim. Behav. Sci., 25: 177-187.

Bateson, P., 1991. Assessment of pain in animals. Anim. Behav., 42: 827-839. Broom, D.M., 1991. Assessing welfare and suffering. Behav. Proc., 25: 117-123. Dawkins, MS., 1980. Animal Suffering. Chapman and Hall, London. Dawkins, MS., 1983. Battery hens name their price: consumer demand theory and the measurement of

ethological ‘needs’. Anim. Behav., 31: 1195-1205. Duncan, I.J.H. and Dawkins, M.S., 1983. The problem of assessing ‘well-being’ and ‘suffering’ in farm

animals. In: D. Smidt (Editor), Indicators Relevant to Farm Animal Welfare. Matinus Nijhoff, The Hague, pp. 13-24.

Fraser, D., 1995. Science, values and animal welfare: exploring the ‘inextricable connection’. Anim. Welfare, 4: 103-l 17.

Friihlich, E. and Oester, H., 1989. Anwendung ethologischer Erkenntnisse bei der Priifung der Tiergerechtheit von Stalleinrichtungen und Haltungssystemen fur Legehennen. In: Aktuelle Arbeiten zur artgemlssen Tierhaltung 1988, KTBL-Schrift 336. KTBL, Darmstadt, pp. 273-284.

Jensen, P., 1989. Nest site choice and nest building of free-ranging domestic pigs due to farrow. Appl. Anim. Behav. Sci., 22: 13-21.

Jensen, P. and Toates, F.M., 1993. Who needs ‘behavioural needs’? Motivational aspects of the needs of animals. Appl. Anim. Behav. Sci., 37: 161-181.

Mason, Cl. and Mendl, M., 1993. Why is there no simple way of measuring animal welfare? Anim. Welfare, 2: 301-319.

B. Wechsler et al. /Applied Animal Behmiour Science 53 (1997) 33-43 43

McFarland, D., 1989. Problems of Animal Behaviour. Longman Scientific and Technical, Harlow, UK. Mendl, M., 1991, Some problems with the concept of a cut-off point for determining when an animal’s welfare

is at risk. Appl. Anim. Behav. Sci., 31: 139-146. Moberg, G.P., 1987. A model for assessing the impact of behavioral stress on domestic animals. J. Anim. Sci..

65: 1228-1235. Oester, H., 1982. Indikatoren fur die Beurteilung der Tiergerechtheit von Haltungssystemen fur Legehennen.

In: Aktuelle Arbeiten zur artgemlssen Tierhaltung 1981, KTBL-Schrift 281. KTBL, Darmstadt. pp. 141-149.

Oester, H. and FrBhlich, E., 1986. Die Beurteilung der Tiergerechtheit der neuen Haltungssysteme fiir Legehennen im Rahmen der Tierschutzgesetzgebung. Schweiz. Arch. Tierheilkd., 128: 521-534.

Oswald. T., 1992. Der Kuhtrainer. Schriftenreihe der Eidg. Forschungsanstalt fir Agrarwirtschaft und Landtechnik, Vol. 37. FAT, Tanikon.

Rushen, J., 1991. Problems associated with the interpretation of physiological data in the assessment of animal welfare. Appl. Anim. Behav. Sci., 28: 381-386.

Ruahen, J. and de Passill&, A.M.B., 1992. The scientific assessment of the impact of housing on animal welfare: a critical review. Can. J. Anim. Sci., 72: 721-743.

Sandoe. P. and Simonsen, H.B., 1992. Assessing animal welfare: where does science end and philosophy begin? Anim. Welfare. 1: 257-267.

Schmid, H., 1992. Arttypische Strukturierung der Abferkelbucht. In: Aktuelle Arbeiten zur artgemassen Tierhaltung 1991, KTBL-Schrift 351. KTBL, Darmstadt. pp. 27-36.

Schmid, H., 1993. Ethological design of a practicable farrowing pen. In: M. Nichelmann, H.K. Wierenga and S. Braun (Editors), Proc. of the Int. Congress on Applied Ethology, Berlin, 1993. KTBL, Darmstadt, pp. 238-242.

Schmid. H. and Weber, R., 1992. Abferkelbuchten: ein neues Konzept. Eidg. Forschungsanstalt fur Agrar- wirtschaft und Landtechnik, FAT-Bericht 417, FAT, Tanikon.

Stoma, A. and Wood-Gush, D.G.M.. 1984. The identification of behavioural key features and their incorpora- tion into a housing design for pigs. Ann. Rech. Vet., 15: 287-298.

Tannenbaum, J., 1991. Ethics and animal welfare: the inextricable connection, J. Am. Vet. Med. Assoc., 19X: I360- 1376.

Troxler. J. and Steiger, A., 1982. lndikatoren fir nicht tiergerechte Haltungsformen in der Schweinehaltung. In: Aktuelle Arbeiten zur artgemassen Tierhaltung 198 I, KTBL-Schrift 281. KTBL. Darmstadt, pp. 150-154.

Troxler. J. and Weber, R., 1989. Anwendung ethologischer Erkenntnisse bei der Priifung von Stalleinrichtun- gen fur Schweine. In: Aktueile Arbeiten zur artgemlssen Tierhaltung 1988, KTBL-Schrift 336. KTBL, Darmstadt, pp. 142-149.

Tschanz, B., 1982. Verhalten, Bedarf und Bedarfsdeckung bei Nutztieren. In: Aktuelle Arbeiten zur artgemassen Tierhaltung 198 1, KTBL-Schrift 28 1. KTBL, Darmstadt, pp. I 14- 128.

Tschanz, B., 1987. Ethology and animal protection, In: D.W. Folsch and A. Nabholz (Editors), Ethical, Ethological and Legal Aspects of Intensive Farm Animal Management. Birkhauser, Basel, pp. 41-47.

Weber. R., 1987. Abferkelbuchten mit Kastenstand. Eidg. Forschungsanstalt fur Agrarwirtschaft und Landtechnik. FAT-Bericht 3 18, FAT, Tanikon.

Weber, R. and Troxler, J.. 1988. Die Bedeutung der Zeitdauer der Geburt in verschiedenen Abferkelbuchten zur Beurteilung auf Tiergerechtheit. In: Aktuelle Arbeiten zur artgemasaen Tierhaltung 1987, KTBL-Schrift 323. KTBL, Darmstadt, pp. 172- 184.

Wechsler. B., 1993. Verhaltensstorungen und Wohlbefinden: ethologische eberlegungen. In: G. Martin (Editor), Leiden und Verhaltensstorungen bei Tieren. Birkhauser, Basel, pp. 50-64.