Adaptation of large herbivores to secondary plant compounds found in their food plants may influence their food choice. Plants, which are potentially toxic, may initially be avoided by herbivores but, following sampling and adaptation of detoxification pathways, herbivores may be able to exploit certain toxic plants without detrimental effects. Thus, adaptation of the rumen microbial population can protect the host animal from the toxic effects of plant secondary compounds. Oxalic acid is a plant component found at high concentrations in a range of food plants, such as Beta vulgaris L, and Rumex spp., commonly consumed by ruminants. It is readily degraded in the rumen by Oxalobacter formigenes following a period of adaptation (Allison and Reddy, 1984).

There is currently no published information on the conditions under which young pigs are transported. European Directive 95/29/EC concerns the protection of animals during transport but recommends no stocking density for pigs moved by road other than those of 100 kg live weight. A survey of 97 farms showed store pigs accounted for 64% of all inter-farm journeys. The most common weights were 26 to 30 kg. The mean stocking density for transport was 162kg/m2 (range 72-214 (s.d. 25) kg/m2). The mean journey time was 165 min and the distance 228 km. This experiment was designed within the range of stocking densities and distances travelled found in current commercial practice. Pigs were transported at high, medium and low densities. Posture and heart rate during transport were recorded.

Although the need for pigs to lie down on long journeys is not in question, there is evidence that they may not choose to do so on journeys of less than 3h (Hunter et al., 1994). These observations were undertaken to determine how pigs in the 95 to 100 kg weight range behaved on short journeys. A three-tier floating-deck vehicle with weld-mesh flooring was used because the popularity of three-deck vehicles is increasing (Riches et al., 1996).

The thermal insulation of pigs (Sus scrofa) is considered poor (Mount and Ingram, 1965) as they do not have a thick coat and have a variable subcutaneous fat thickness. Ventilation rates in animal houses are often reduced in colder weather to maintain an optimal thermal environment. However this may fail to remove atmospheric pollutants (e.g. Menses, 1987).

The concept of ‘coping strategies’ has received increasing attention over the last few years. Benus et al. (1991) reported two major types of strategy in rodents. These were the so-called active and passive strategies. The active strategy consists of being aggressive, forming behavioural routines and showing low responsiveness to changes in the environment, the passive strategy involves being low aggressive, flexible and responsive to changes in the environment. Hessing et al. (1993) reported similar findings in pigs. These strategies have potential implications for pig husbandry, since they may be more or less adaptive in specific farming environments.

Testing several genotypes on one diet may impose nutritional constraints on growth of one or more of those genotypes. Use of a diet choice procedure may enable pigs to ‘choose’ the appropriate combination of the diets to attain the growth determined by their genotype. In an animal breeding context, the diet choice procedure offers the potential to increase the accuracy of predicting an animal’s genetic merit for performance test traits. In the few diet choice studies made with pigs heavier than 35 kg (e.g. Bradford and Gous, 1991), the animals have been of one genotype. Therefore, animals from the Edinburgh lean growth selection experiment (Cameron, 1994) provided the experimental resource to determine if between-selection line differences were enhanced by testing pigs on a diet choice procedure compared with testing on a single diet.

The practice of early weaning of young piglets is associated with a sudden change in both their diet and their method of food acquisition. It can be argued that 4-week-old piglets are neither physiologically nor behaviourally prepared for these rapid changes and the provision of solid creep food is largely ineffective in facilitating adaption to weaning (Fraser et al., 1994). Previous studies which have attempted to promote solid food intake in pre-weaned piglets have tended to focus on dietary factors, such as palatability and digestibility (Bruneau and Chavez, 1995). However for ingestion to occur piglets must first perceive the food items and then establish an attraction to and acceptance of them.

Young animals learn feeding through trial and error processes and from social models, especially the parental one (i.e. Suboski and Bartshumas, 1984). Under practical conditions the mother-young bond has been reduced by early weaning in piglets or deleted in artificially hatched chicks. In both cases feeding experiences develop in groups of congeners of the same age. Negative reinforcements of feeding experiences have been demonstrated with diets containing concanavalin A from jackbean seeds (Canavalia ensiformis) (Leon et al., 1991). Preliminary trials suggested that piglets react similarly to jackbean containing diets. The aim of the present experiment was to investigate the capacity of piglets and chicks to influence the feeding behaviour of conspecifics according to their previous feeding experience (negative with concanavalin A, or neutral) and their social reactivity.

Broiler chickens have been selected for increased growth rate and adapted to consumer demands. A range of commercial products is being developed from slower growing country-type meat chickens (named ‘Label’ in France) as well as from faster growing broilers with a high yield of breast meat. Nutritionists have to satisfy a need for feeding programmes adapted to meet the demands for various end-products. Selection for growth rate has to be shown to change the food intake behaviour of chickens (i.e. Barbato et al., 1980). Emmans (1991) demonstrated that chickens are able to adjust their food choices if the paradigm permits an adequate choice. In a series of experiments (see Picard et al., 1994 for review) the authors concluded that ‘high producing animals are not necessarily those that will react most clearly to an amino acid deficiency by altering their food intake and/or their feeding behavior’.

Chickens have been selected for millenniums on their ability to select food in complex and variable environments. Artificial selection for juvenile body weight using a single balanced food might have modified the ability of chickens to adapt to a choice feeding situation (Siegel and Dunnington, 1990). However, diet selection for protein has been demonstrated in many recently published experiments (for review: Forbes and Shariatmadari, 1994).

Birds attempt to compensate for an E : P (energy: protein) imbalance by increasing their intake of protein or energy from selected foods and this is related to changes in the birds’ body composition caused by corticosterone treatment.

Corticosterone injections increase fatness despite increasing nitrogen and energy excretion (Bartov, 1985). More recently corticosterone has been used to modulate macronutrient selection in rats (Devenport et al., 1991; Bligh et al., 1993; Tempel et al., 1993) and chickens (Covasa and Forbes, 1995). The intake of high protein concentrate was decreased by corticosterone treatment in young chickens but not in older chickens (Covasa and Forbes, 1995). The previous results suggest that more information on corticosterone’s effect on diet selection will be obtained by using foods which can offer a wide range of choices to compose a proper diet (Covasa and Forbes, 1995). Therefore, two experiments were conducted to detect changes in protein preference in relation to reduced protein deposition and enhanced fattening induced by corticosterone in male and female growing broiler chickens.

Leg weakness reduces the amount of time broilers spend on their feet (Weeks and Kestin, 1994). It has been shown for wild birds that the frequency of visits to a food source relate to the cost of reaching it (Kacelnik, 1984). We recently reported (Weeks and Davies, 1996) that this appeared to apply to lame broilers which fed for approximately twice as long but half as frequently as their mobile contemporaries.

A large proportion of broiler chickens suffer from lameness euphemistically called ‘leg weakness’. In a survey of commercial, intensively reared broilers, 90% had a detectable gait abnormality and 26% suffered an abnormality of sufficient severity for their welfare to be compromised (Kestin et al., 1992). It is assumed that leg weakness is painful but there is little direct evidence of this. Chickens have been shown to be able to select an adequate protein diet from a choice of two or three foods which are individually inadequate (Forbes and Shariatmadari, 1994). Broiler chickens showed a significant preference for food of the colour which was paired with ascorbic acid supplementation when the requirement for ascorbic acid was increased by heat stress (Kutlu and Forbes, 1993).

Captive animals may suffer if they cannot perform important behaviour and changes to their housing will be most effective at improving welfare if the modifications provide opportunities to perform those behaviours which are most important to the animal concerned (Dawkins, 1983). One means of identifying these behavioural priorities is to place a cost on their performance, as animals persist with the performance of the most important behaviours in the face of rising cost but cease to perform less important behaviours (Dawkins, 1983). There is, however, considerable debate over how best to measure persistence or elasticity of demand (Dawkins, 1990; Mason et al., 1997a) in terms of maximizing both their internal and external validity. To produce results with maximal internal validity it is important to follow the guidelines for using consumer demand theory presented by Mason et al., (1997b): that animals are tested in closed economies; that they are free to control their behavioural bout lengths; that resource use and cost must covary; and that the animals perceive the imposed cost to be equivalent for all resources.