In this issue of Animal Welfare for the first time we have included an Invited Essay. These essays are to be occasional features on contemporary welfare problems; written at the invitation of the editor and refereed. The first is by Trevor Poole, a mammal behaviour zoologist with an active interest in laboratory animal science and welfare, considering the new and controversial subject of transgenic animals. In this fast moving field relatively little thought has been given to the possible suffering which transgenic animals may be enduring for human benefit.

When this journal was first produced in January 1992, the intent was to publish scientific and technical papers that would have a direct benefit on the welfare of animals kept on farms, in laboratories, as companions, in zoos or as managed by man in the wild. The submitted Articles in this fourth part of the fourth volume certainly span man's use of animals and they also, I believe, demonstrate that the knowledge and understanding gained during technical studies can often be used to improve the well-being of the animals themselves.

The first contribution in this issue of Animal Welfare is the second of the newly instigated Invited Essays, written at the invitation of the editor and aimed at addressing contemporary welfare problems. It is by Robert Hubrecht, Assistant Director of UFAW and it examines the techniques involved in the genetic modification of animals, and considers the welfare implications and the relevance of the current UK legislation. These essays, which will be an occasional feature of the journal, are refereed by independent experts but very much represent the personal views of the authors. Hopefully some of the opinions expressed will stimulate readers to respond with constructive critical comments, which could be published in the journal's Letters section. The original author will of course have the right to reply.

The main intent of the Joumal has been to provide a place of publication for papers covering technical aspects of animal welfare and, in particular, to publish accounts of unique field material which might not otherwise be formally reported.

Introduction

The first experiments in genetic engineering took place in the late sixties and early seventies, but animal transgenics (the insertion of DNA into the genome of an animal) really came of age with the production of mice whose genetic sequences contained a gene for human growth hormone (Palmiter et al 1982). Since then there has been a dramatic increase in the quantity of research into genetic modification, using various techniques. This has been largely driven by the potential medical and financial benefits of being able to modify animal and plant genomes (Carver et al 1993; Smith 1994; Stewart et al 1993).

Transgenic animals are becoming increasingly important in laboratory animal science. They seem to offer many potential benefits for mankind particularly in the fields of immunology, teratology and our understanding of the role of specific genes in development. The advantages of using transgenic animals could be particularly important in disease control and in the production of biologically active substances. It may also be possible, by inducing a pathological condition simulating a human disease in, for example, a mouse, to avoid using a mammal higher in the phylogenetic scale which, through its closer relationship with man, might otherwise be the only available species showing the condition.

Zebra finches are often housed in monogamous pairs in small barren cages. In the present study it was investigated whether modest enrichment resulted in behavioural changes. Four plywood box cages of two different lengths were compared. A small and a large cage were used for the barren environment. Another small and large cage were used for the enriched environment, and included additional perches, twigs as well as sand and water baths. An ethogram consisting of 19 different behavioural patterns was used. The behaviour of the birds was studied using a combination of alternating (20s) one-zero and instantaneous sampling, one hour per day during a continuous four-week period.

Locomotor activity, vocalization and singing was significantly increased in the enriched cages and flying was more frequent in the large enriched cage than in any of the other cages. The study demonstrated the usefulness of behavioural analyses to determine to what extent simple enrichment allows the birds to exhibit a wider range of their natural behaviour.

The aversive effects of 90 per cent argon in air, 30 per cent carbon dioxide in air or 90 per cent carbon dioxide in air were investigated in slaughter weight pigs. Aversion was assessed from their reluctance to enter the three gaseous atmospheres to obtain a reward (apples). The pigs did not show any aversion to the inhalation of 90 per cent argon in air. The majority of the pigs did not show aversion to the presence of 30 per cent carbon dioxide in air. By contrast, the inhalation of 90 per cent carbon dioxide was aversive to the majority of the pigs. Fasting them for up to 24h prior to testing did not overcome the pigs ‘ reluctance to enter an atmosphere containing 90 per cent carbon dioxide.

During capture, in order to separate him from a possessive adult female and return him to his mother, a newborn male in a laboratory group of Cebus capucinus monkeys was found to have a seriously infected compound fracture of the humerus associated with a deep and extensive slash wound. Amputation of the affected limb was deemed necessary. Shortly after surgery the newborn was returned to his mother, in isolation from the group, with periodic removal for post-surgical care. Three weeks later the mother-newborn pair was returned to the social group and no further intervention occurred. Regular observations revealed mutual behavioural adjustments to the handicap by the mother and newborn. Compared to a normal age-mate, the amputee received more positive social attention from the mother and other group-members. Despite his showing delays in locomotor and manipulatory activities, the handicapped infant showed good behavioural progress. Early resocialization thus appears feasible following emergency surgery in newborn primates.

The purpose of this study was to investigate the behavioural effects of keeping circus elephants in paddocks. Therefore some species-typical behavioural characters and the occurrence of stereotyped movements were observed and compared in unshackled paddock keeping versus shackled keeping. The investigation includes 29 elephants (19 Elephas maximus, 10 Loxodonta africana) at four circuses located in Germany and in Switzerland.

The results showed that paddocks offered more freedom for comfort, social and play behaviour since such activities were observed more frequently in paddocks than in shackled keeping. Also, stereotyped movements were nearly absent in paddocks and very frequent in shackled keeping. In comparison to shackled keeping, paddocks were more suitable for the needs of elephants.

Recently, the method by which Japanese and Norwegian whalers kill minke whales has come under scrutiny on welfare grounds. Whales are still caught with a grenade tipped harpoon fired from a cannon. The harpoon is targeted to strike the animal in the thorax, though the pattern of harpoon strikes is variable and some can even be struck in the tail. Once the whale is harpooned, it is winched to the ship and if not dead, attempts are made to kill it either by electrocution or with rifles. Some studies indicate that less than 30 per cent of animals are killed instantaneously, though best practice can achieve 50 per cent of animals killed instantaneously but with wounded animals surviving for up to one hour. To kill any wild animal humanely, immediate insensibility must be induced. Whilst in principle the methods used on whales could induce immediate insensibility, there are indications that neither harpooning nor the secondary killing processes, like electrocution or rifle bullets, are achieving this in an acceptable proportion of the animals taken. Comparison of the killing processes used on minke whales with killing processes used on other animals, indicated that there are several areas where improvements could increase the proportion of whales killed instantaneously and reduce the suffering of wounded animals. Whilst harpooning may remain the favoured method of taking whales, and some improvements have been made in the number of whales killed instantaneously, the equipment has evolved little from that originally developed by the Norwegians in the Nineteenth Century. As currently practised few people would consider the current methods employed by the whaling industry to be humane.

In Britain large numbers of animals are taken into captivity for treatment or care and then subsequently returned to the wild, but there are few data on the effectiveness of these rehabilitation programmes. In this study, over a period of four years 251 fox cubs that had been captive-reared were tagged and released; 90 were recovered. Survival rates were low, and road traffic accidents were found to be a major cause of mortality immediately following release. Recovery distances were lower than expected. The stress associated with captive-rearing meant that released foxes weighed less than wild-reared foxes, and they suffered further weight loss in the period immediately following release, even though an analysis of the stomach contents of animals recovered dead showed that released foxes rapidly learnt to hunt successfully.

It was concluded that captive-rearing is a problematic process for foxes, and contrary to predictions they face severe problems in adapting following release. Suggestions are made for the improvement of fox captive-rearing and release programmes, and the need for similar studies on other species is highlighted.

In conceptualizing animal welfare, it is useful to distinguish among three types of concepts. ‘Type 1” are single, measurable attributes. ‘Type 2’ are single attributes that cannot be measured directly but can be estimated by correctly combining various contributing attributes. ‘Type 3’ are concepts involving multiple attributes which are grouped together because they serve some common function, and whose relative importance cannot be established in an entirely objective way. Individuals who treat animal welfare as a type 1 concept may propose single, objective measures of welfare, such as longevity or levels of stress-related hormones; however, this approach rests on judgements, which are not purely objective, about the relative importance of different factors for an animal's quality of life. Studies of animal preferences and motivation are sometimes seen as an objective way to weight different attributes according to the animals’ own priorities, and thus allow animal welfare to be treated as a type 2 concept. However, numerous technical and fundamental difficulties limit our ability to do this. Animal welfare is best seen as a type 3 concept incorporating multiple attributes, with considerable consensus over certain general principles (eg that a high level of welfare implies freedom from suffering) but with value-related disagreement over how these principles should be applied. Because the various attributes cannot be combined in a purely objective way, science is limited in its ability to determine the ‘overall’ welfare of an animal and to compare welfare in disparate environments. Instead of attempting to ‘measure’ animal welfare, the role of science should be seen as identifying, rectifying and preventing welfare problems.

As part of a study into the effects of human activities on the welfare of free-living wildlife, the relative scale and severity of welfare problems in wild mammals and birds in Europe were investigated. Major cases were described and compared in terms of the nature and level of harm (pain, stress and fear) they cause, the duration of these effects and the number of individuals affected. The use of anticoagulant rodenticides, myxomatosis in rabbits, the poisoning of wildfowl by ingested lead shot, the contamination of seabirds with fuel oil, the effects of shooting, injuries due to collisions with road traffic and prédation by domestic cats all severely compromise the welfare of large numbers of animals. Practical approaches to the alleviation and prevention of some of these welfare problems are discussed. We suggest that in assessing the environmental impact of new developments and technologies prior to their implementation, possible consequences to wildlife welfare should always be considered.

A haulage company, transporting pigs for a marketing group, has monitored information about journeys and kept records of deaths in transit (DIT). These data can be used to set guidelines and they highlight those areas of pig transport that need further study.

Factors, other than genetic ones, which may contribute to the rate of DIT were considered. A strict seasonal effect was not apparent, but where conditions were hot and wet there was an increase in DIT. There was also a ‘time of day” effect, with transport in the afternoon showing a higher rate of DIT than transport in the morning. It was observed that if pigs were deemed dirty at loading then DIT were more likely. Clean pigs at loading could indicate that hygienic and environmental conditions in the finishing stage were good. Individual producers showed a wide variation in their rate of DIT, with some producers regularly experiencing a high level Routine monitoring of the rate of DIT and transport conditions could help to pinpoint the source, and possible cause, of the problem.

Information that helps in the understanding of transit deaths is important because if stress in the period from selection to slaughter can be reduced, there should be both economic as well as welfare benefits.

The release of animals from captivity frequently leads to a period of erratic movement behaviour which is thought to expose the animal to a high risk of mortality. Twenty-six foxes which had been reared at a wildlife hospital or captive-bred, were radio-collared when nearly full-grown and released without site acclimation. Immediately after release there was an erratic phase of behaviour, during which the foxes travelled widely and movement parameters were markedly elevated. For those foxes which survived, a second phase was entered after an average of 17.2 days, during which one small area only was used, and movement parameters were much reduced. In a second study, nine foxes were released following site acclimation in a pre-release pen; this process postponed but did not eliminate the phase of high movement activity.

This pattern of movement was compared with the dispersal behaviour of wild-reared foxes. It was concluded that released foxes, despite being proficient in other aspects of behaviour, were moving and behaving in a markedly abnormal manner and this resulted in a high death rate. The results are used to discuss methods of improving rehabilitation techniques.

This study examined the effects of enriching the environment on the behaviour and productivity of pigs, in an attempt to assess changes in welfare. A total of 102 pigs were housed in either barren or enriched housing from birth to slaughter. The barren environments were defined as intensive housing and the enriched environments incorporated extra space, an area which contained peat and straw in a rack. Enriching the environment reduced both the amount of time pigs spent inactive and the time involved in harmful social and aggressive behaviour. Tail biting was absent from the enriched environment but four pigs were removed from barren pens with severe tail damage. Pigs housed in enriched environments spent longer durations in exploratory behaviour than those in barren housing, and young pigs in enriched environments performed locomotory behaviour more frequently than their counterparts in barren environments. Overall growth rates were similar in both treatments. These results indicate that welfare is improved by enrichment with substrates and suggest that barren pens should be modified to provide these facilities.

The welfare of caged laying hens could be improved by placing objects in the feed trough. Such objects might (a) simulate general ground-litter thus promoting more normal foraging activity and (b) give hens the opportunity to work ‘for feed - a behaviour usually thwarted in conventional cages. Spherical objects with various characteristics were placed in the feed trough of a tier of caged laying hens (n = 16). The hens pecked frequently at the objects, moving them to the trough space of adjacent cages. The mean proportion of hen heads over the trough containing these objects was significantly greater than before the objects were present (35.3 cf 32.9%) and significantly greater than the proportion of heads over a similar trough containing no objects (33.6%). Thirty days later, the mean proportions were still significantly different (33.5 cf 31.0%) showing that there was little habituation. Daily manual scattering of the objects increased the distance they were subsequently moved by the hens (23.0 cf 19.3 cm/day) indicating increased pecking activity. In a second study 12 hens were given a choice of feeding from troughs containing 0, 12 or 36 spherical objects. There was no overall preference to feed from any of the troughs. All the hens fed from troughs containing the objects, possibly indicating that the opportunity to move the objects and forage or work for feed was desired on occasions. Brightly coloured spherical objects are considered to be a promising method of successful environmental enrichment for caged laying hens. Their use to improve the welfare of caged laying hens appears to be practical and reasonably inexpensive.

A test system previously described by Blom et al (1993) was used to investigate the preference of the Mongolian gerbil (Meriones unguiculatus) with respect to light intensity in their cage, by comparing a standard transparent rodent cage with three cages darkened by either 25, 50 or 75 per cent. The ten animals included in the test spent most of their time in the cages that were darkened by 50 and 75 per cent, thus relatively avoiding the standard cage. Preference was mainly determined by visits longer than 15min and was therefore possibly associated with sleeping behaviour. Food and water consumption were not restricted to the preferred cages.

These results imply that gerbils should be housed preferably in cages that allow sheltering in relatively dark places. For this purpose, placing shelter objects in the cage should be considered.

A review of the scientific literature gives evidence that transferring previously single-caged adult macaques to permanent compatible pair-housing arrangements (isosexual pairs, adult/infant pairs) is associated with less risk of injury and morbidity than transferring them to permanent group-housing arrangements. Juvenile animals can readily be transferred to permanent group-housing situations without undue risks. Safe pair formation and subsequent pair-housing techniques have been developed for female and male rhesus (Macaca mulatta), stump-tailed (M. arctoides) and pig-tailed macaques (M. nemestrina) as well as for female long-tailed macaques (M. fascicularis). Pair housing does not jeopardize the animals’ physical health but it increases their behavioural health by providing them with an adequate environment to satisfy their need for social contact and social interaction.