Cultivated pastures in southern China are being used to improve forage productivity and animal performance, but studies on grazing behaviour of goats in these cultivated pastures are still rare. In the current study, the grazing behaviour of Yunling black goats under low (5 goats/ha) and high (15 goats/ha) stocking rates (SRs) was evaluated. Data showed that the proportion of time goats spent on activities was: eating (0.59–0.87), ruminating (0.05–0.35), walking (0.03–0.06) and resting (0.01–0.03). Compared with low SR, goats spent more time eating and walking, and less time ruminating and resting under high SR. Goats had similar diet preferences under both SR and preferred to eat grasses (ryegrass and cocksfoot) more than a legume (white clover). The distribution of eating time on each forage species was more uniform under high v. low SR. Bites/step, bite weight and daily intake were greater under low than high SR. Results suggest that the SR affects grazing behaviour of goats on cultivated pasture, and identifying an optimal SR is critical for increasing bite weight and intake.

Herbivore foraging is influenced by spatial and seasonal changes in the production of leaves and fruits. To understand how herbivores respond to these changes, it is necessary to identify their habitat preferences and how they use the vegetation available. In the Neotropical region, one of the largest terrestrial herbivores is the lowland tapir (Tapirus terrestris), a species important for its dual role as browser and seed disperser. The objective of this study was to determine the species and plant components (leaves, fruits) utilized by T. terrestris in different time periods and habitats, in relation to changes in food availability in the north-western Amazon. Tapir diet was established through identification of browsing signs and faecal analysis, from data collected in the field during the months of March, April, August and September of 2015. Plant species availability for browsing was sampled in ten 2 × 50-m transects and fruit productivity was estimated in linear transects (~9 km). We found that T. terrestris mostly consumed vegetative parts, i.e. fibre (70–90%), and to a lesser extent fruits (10–30%). Food consumption was selective and concentrated in habitats with higher availability of preferred plants. When fruit intake increased, selectivity in browsing became more intense and limited to preferred species. This information, coupled with our findings about seasonal differences in browsing vs. frugivory patterns, provides valuable knowledge for understanding how environmental heterogeneity may influence the foraging ecology of the lowland tapir.

The number of horses in northern Spanish mountains has increased in recent decades, but little is known about their grazing behaviour, performance and potential for foal meat production. This research aimed to study the diet selection, liveweight (LW) changes and parasitic status of dry and lactating mares, and foals’ LW gains, grazing on heathlands with different botanical composition. The experimental design consisted of three vegetation types: dominated by heather (Ericaceae) species (H), dominated by gorse (Ulex gallii; G) and co-dominated by gorse and heath-grasses (G-G), with four replicates per treatment (12 paddocks of 1.2 ha). The study lasted three grazing seasons (2010–12). Each year, 24 crossbred mature mares (310±52 kg LW) were used, managing one lactating mare with her foal plus one non-lactating mare per paddock from May to late summer or early autumn. In the case of H paddocks, animals had to be removed before (late August to early September) because of apparent loss of body condition. Animals were periodically weighed. Mares’ diet composition was estimated using alkane markers, analysing the discrepancies in alkane concentrations between dietary plant components and faeces. Faecal samples were also analysed for gastrointestinal nematodes ova. Chemical composition of the main plant components (i.e. heather, gorse and grasses) revealed a low nutritive value, averaging 79, 115 and 113 g CP/kg dry matter (DM), respectively, that could restrict livestock performance. Mares initially selected gorse and grasses (0.47 and 0.40, respectively, in 2010), increasing heather consumption over time (from 0.13 in 2010 to 0.29 in 2012) as gorse availability decreased. The performance of both mares and foals was lower in H compared with G and G-G paddocks (−216 v. 347 g/day for mares, P<0.01; 278 v. 576 g/day for foals, P<0.05), whereas LW changes were more favourable in dry mares than in lactating ones (241 v. 78 g/day; P<0.05). Small strongyle (Cyathostominae) egg counts in mares’ faeces increased across the grazing season with no differences between treatments. These results indicate that grazing by horses on gorse- and grass-gorse-dominated shrublands could be sustainable at least during part of the year (4 to 6 months). However, heather-dominated heathlands are not able to meet the nutritional needs of horses even for a short time (2 to 4 months). Nevertheless, the low nutritive quality of these vegetation communities, especially in autumn, requires animal access to other pastures with a higher nutritive value, or supplementary feeding, to enhance foals’ growth and maintain sustainable grazing systems with productive herds.

Past positive experiences can increase herbivores’ motivation to eat low-quality foods. However, this is not always translated into a higher preference for low-quality foods in choice tests among foods of higher nutritional quality. Foraging behavior is also affected by properties of the feeding context because the quality and abundance of foods in nature change in time and space. We hypothesized that in a choice situation, the expression of a past positive experience with a low-quality food is modulated by the costs associated with selecting a high-quality food option. A total of 24 sheep were randomly assigned into two groups (n=12). During conditioning phase, one group (CS+; i.e., conditioned group) was fed with oat hay (a low-quality food) for 20 min and immediately after a ration of soybean meal (a nutritious food), whereas the other group was also fed with oat hay but the offer of soybean meal was delayed 5 h (CS−; i.e., control group). After conditioning, we assessed sheep motivation to eat the oat hay in an experimental arena in which accessibility to alfalfa hay (a high-quality food) was increasingly restricted. When alfalfa hay was readily accessible, CS+ and CS− sheep almost exclusively selected this food, showing a small and similar preference for oat hay. However, when accessibility to alfalfa hay decreased, intake and selection of oat hay was greater in the CS+ sheep than in the CS− sheep. The latter was a consequence of differential changes in behavior between groups; for example, sheep in CS+ spent more time foraging oat hay and were more likely to switch to oat hay if they had previously been eating alfalfa hay than sheep in CS−. Our results show that behavioral expression of the conditioned preference for a low-quality food depends on parameters of the feeding context (e.g., availability). We suggest that this can be the link between learning models and optimal foraging models of diet selection.

Two experiments were conducted to assess preference by sheep for endophyte-infected tall fescue growing in monoculture at least 5 m away from alfalfa (fescue-middle (FM)) over endophyte-infected tall fescue growing adjacent (0.2 to 1 m; fescue–alfalfa (FA)) to alfalfa (FA), and the effect of legume scent on preference for endophyte-infected tall fescue. In Experiment 1, 10 six-month-old lambs were offered for 12 days a choice of freshly harvested FA and FM. On days 13 and 14, lambs were offered the same choice, except cages (to allow access only to scent) containing freshly harvested alfalfa were put in the feeders containing FA, whereas cages containing freshly harvested FM were included with the feeders containing FM. Forage intake was measured 1 h after feeding and at three consecutive 2-h intervals thereafter. FA contained greater (P<0.002) concentrations of the alkaloid ergovaline (360±27 ppm) and CP (8±0.4%) than FM (219±27 ppm and 6±0.4%, respectively). Lambs preferred (P<0.05) FA to FM during the 1st hour of feeding, but the differences became smaller and disappeared in later feeding periods (P<0.005). Lambs offered FA with alfalfa scent or FM with FM scent preferred (P<0.05) FA but only on the 2nd day. In Experiment 2, 10 six-month-old lambs were offered a choice of FM with cages (to allow access only to scent) containing freshly harvested alfalfa or FM for 8 days. During the following 4 days, FM in the cages was replaced with freshly harvested sainfoin. Preference was greater (P<0.05) for FM offered with alfalfa scent than for FM offered with FM scent only on days 4 and 8. When lambs were offered FM with alfalfa or sainfoin in cages, they preferred (P<0.05) tall fescue with sainfoin scent over fescue with alfalfa scent, but intake was variable across hours and days (P<0.001). It is concluded that (1) lambs adjusted their intake of and preference for FA and FM over successive feeding bouts within each day, likely owing to an attempt to balance intakes of nutrients and alkaloids and (2) olfactory cues influenced preference, but to a lesser extent than nutrients and alkaloids in endophyte-infected tall fescue.

We conducted an experiment to determine whether early-life social learning of feeding site selection in lambs was sex-specific. Sixteen ewes and their new born lambs were used in a controlled experiment. Eight ewe–lamb pairs included a male lamb and the remaining eight a female lamb. All pairs were individually exposed to an experimental arena containing a safe and unsafe artificial feeding site (SFS, UFS) each consisting of nine bowls which contained either ground Bermuda grass hay (SFS) or ground alfalfa hay (UFS). The bowls in UFS were surrounded by bright orange traffic cones (visual cues). Half the ewes were trained with controlled electric shock to avoid UFS. Thus, pairs were randomly assigned to: (1) shock aversion training (SAT) to mothers of male lambs (MS); (2) SAT to mothers of female lambs (FS); (3) no aversion training (NAT, control) to mothers of male lambs (MC); and (4) NAT (control) to mothers of female lambs (FC). None of the lambs were subjected to SAT. During training, testing, extinction, and retraining ewe–lamb pairs were exposed to the arena together. Ewes were then removed from the experiment and two additional extinction phases were conducted with weaned lambs alone. Fear conditioning elicited UFS avoidance of both the trained ewes (means±s.e.m. % times observed in UFS during testing phase: FC=95.3±1.70; MC=94.4±4.87; FS=1.6±1.63; MS=0 ±0; P<0.01) and their naïve lambs (FC=83.8±6.07%; MC=76.6±6.56%; FS=30.4±7.90%; MS=33.9±9.23%; P<0.01). UFS avoidance in lambs occurred regardless of sex and tended to persist after weaning (% times observed in UFS during 1st post-weaning extinction phase: FC=92.6±4.50%; MC=89.8±6.09%; FS=45.1±10.57%; MS=43.5±10.42%; P=0.06). Fear conditioning in mothers appeared to alter sex-related differences in mother–infant behavioral synchrony by increasing and decreasing feeding synchrony of male and female lambs, respectively (FC: r=0.52, P<0.01; MC: r=−0.02, P=0.86; FS: r=0.14, P=0.26; MS: r=0.46, P<0.01). During the extinction phase mothers of ram lambs were observed feeding more often (FC=85.0±2.33%; MC=92.7±1.45%; FS=47.3±8.81%; MS=72±5.68%; P=0.02) and standing less often than ewes with daughters (FC=7.3±2.40%; MC=2.7±0.83%; FS=39.3±9.04%; MS=18.0±5.29%; P=0.06). This study suggests that social conditioning at an early age could be a viable tool to induce learning of feeding site avoidance in female and male lambs alike.

As the number of horses is steadily increasing in Cantabrian less-favoured areas, mostly occupied by nutritionally poor heathland communities, the objective of this research was to study and compare diet selection and performance of crossbred horses and beef cattle (Asturiana de los Valles breed) grazing these areas during the summer. A total of 20 cows (seven lactating their calves and thirteen dry) and 20 mares (eight lactating their foals and twelve dry) were managed during 3 years on a heathland area from June to September and October. The proportion of grass (mostly Pseudarrhenatherum longifolium) and heather (Erica umbellata, E. cinerea and Calluna vulgaris) in the diet selected by cows and mares was estimated in two occasions (July and September) each year using the alkane markers. All animals were weighed at the beginning, middle and end of the summer grazing season. Mares selected heath grasses and rejected heather species more than did the cows (P < 0.001), particularly in early summer (0.85 v. 0.65 herbage proportion in July; 0.63 v. 0.55 in September). Lactating state did not affect diet composition. Mares achieved more favourable (P < 0.001) body weight (BW) changes than cows, but there was also an interaction with season (P < 0.001). From June to August, cows lost BW, being the daily changes (in percentage of initial BW) −0.15%/day in lactating and −0.05%/day in dry cows, whereas both lactating and dry mares gained BW (0.16 and 0.12%/day, respectively). From August to October, lactating cows and mares lost more BW (P < 0.01) compared with their non-lactating counterparts (−0.19 v. −0.10%/day in cows, −0.14 v. −0.05%/day in mares). With regard to the offspring, BW gains between June and August were similar in calves and foals (0.83 and 0.90%/day, respectively), but thereafter calves achieved higher (P < 0.05) BW gains than foals (0.37 v. 0.16%/day). It is concluded that the low nutritive quality of these heathlands restricts, both cattle and horse performance, even during short summer grazing season, being more penalized lactating animals, especially the cows. Although calves were able to maintain acceptable growth rates, foals showed reduced ones during the second half of the grazing season. Other plant communities with better nutritive quality should be provided to assure replenishment of body reserves in dams, and enhance offspring gains to maintain sustainable grazing systems with productive herds of cattle and horses in these heathland areas.

As Rhagodia preissii had shown significant in vitro anthelmintic activity in a previous study, we examined the effect of including this shrub in the diet of sheep infected with Trichostrongylus colubriformis. Worm-infected merino wethers were grazed for 7 weeks on either R. preissii or annual pasture, and faecal egg counts (FECs) were conducted weekly. Plant material was collected weekly from eaten and uneaten plants, and analysed for levels of plant secondary metabolites (tannins, oxalates, saponins) and in vitro anthelmintic activity. While mean FECs were consistently lower in sheep grazing R. preissii compared to pasture (reductions of 20–74%), the differences were not significant. There was no relationship between grazing preference (eaten or uneaten) and in vitro anthelmintic activity of plant extracts. The levels of saponins and oxalates did not correlate with grazing preference or in vitro anthelmintic activity, while tannins were not responsible for the anthelmintic activity. While the identity of the grazing deterrent and in vitro anthelmintic compounds remain unknown, the presence of plants which were both highly preferred by the sheep and showed in vitro anthelmintic activity indicates a potential to develop the species as an anthelmintic shrub through selection of shrub populations dominated by such plants.

Social learning from peers can trigger herd-wide intoxication with white locoweed (Oxytropis sericea), an alkaloid-synthesizing herbaceous legume that grows on rangelands of western North America. We conducted an experiment to test the hypothesis that restriction of the area allocated to animals to feed in would inhibit social facilitation of locoweed ingestion in yearling heifers. Eight heifers that avoided white locoweed (LA) and eight heifers that readily consumed it (LE) were selected from a pool of 40 cross-bred heifers and were randomly assigned to the social facilitation or social interference treatment groups. We conducted 200 10-min feeding trials in three 5-day phases (pre-treatment, treatment, post-treatment) during which animals were presented with a set of bowls arrayed in a test arena, some of which contained ground wheat straw and others contained air-dried ground white locoweed. During the pre-treatment (days 1 to 5) and the post-treatment phases (days 11 to 15) non-social trials were conducted in which the feeding behavior of individual animals was investigated in an 80 m2 arena containing 12 feeding bowls. During the treatment phase (days 6 to 10) social learning trials were conducted in which LA + LE pairs from the social interference group were exposed to 12 bowls of food distributed in an 80 m2 arena intended to induce social interference, and LA + LE pairs from the social facilitation group were exposed to 36 bowls of food distributed in a 240 m2 arena intended to permit social facilitation. During pre-treatment phase, LA heifers consumed detectably less locoweed and wheat straw and exhibited lower preference for locoweed than LE (P ⩽ 0.05) although wheat straw preference of LA and LE was similar. During social learning trials (treatment phase), LA in the social interference group visited similar number of locoweed bowls (mean ± s.e.m.: 0.2 ± 0.12) as they had during non-social learning (0.2 ± 0.20). Conversely, LA heifers in the social facilitation group visited detectably more locoweed bowls during social learning trials (1.6 ± 0.46) compared with the pre-treatment phase (0.2 ± 0.16). Correlation between daily number of locoweed bowls visited by LA and LE during social learning trials was detected in the social facilitation (r = 0.70; P < 0.01), but not in the social interference group (r = 0.15; P = 0.52). During testing trials (post-treatment phase), locoweed and wheat straw intake and preference of LA and LE in both treatment groups was similar. Manipulation of the feeding environment delayed, but did not inhibit social learning of toxic weed ingestion in this study.

Optimal foraging predicts that animals should be able to assess the content of important nutrients in food. Ungulates discriminate salt and P, but discrimination of other minerals is controversial even though they are also essential and often limiting. Animal scientists have explained this taste through palatability, which predicts the same pattern of discrimination for calves and hinds and greater consumption by the latter. Social learning may also be involved, predicting a correlation between mother and calf and less consumption by the latter. The present study examines the consumption behaviour of free-choice supplemented minerals by hinds and calves of Iberian red deer (Cervus elaphus hispanicus) to discern between these hypotheses. Behavioural indices of intake correlated with actual mineral consumption (P < 0·001). Mother and calf behavioural indices correlated only for salt-mixed minerals. Calves showed overall behavioural indices of consumption greater than hinds (P < 0·01 and P < 0·001), and also for all single supplements except NaCl, as expected from growth needs and in contrast to the palatability hypothesis. Calves showed a greater consumption of CuSO4 and lower of Na2SeO3 than pure salt. Hinds showed a different pattern, ingesting lower amounts of all minerals except CuSO4 and salt. Additional analyses also showed discrimination between minerals unmixed with salt, such as CaHPO4 and CaCO3 (P = 0·012 and P = 0·020). The greater intake of growing calves and the different consumption patterns for hinds and calves suggest that deer can discriminate among minerals, and that they do not consume minerals for their palatability or driven by social learning. Therefore, deer may be selecting minerals according to nutritional requirements.

This study examined the effects of physical form and stage of maturity at harvest of whole-crop barley silage (WCBS) on feed intake, eating and rumination activity, diet selection and faecal particle size in dairy steers. Whole-crop barley was harvested and ensiled in round bales. Eight dairy steers (live weight (LW): 350 ± 10 kg) in a duplicated 4 × 4 Latin square design were fed WCBS harvested at heading or dough stage of maturity in long form or chopped in a 2 × 2 factorial treatment arrangement. The WCBS was supplemented with soybean meal. Daily dry matter (DM) intake increased by 7% (P < 0.05) due to chopping of dough-stage silage but was unaffected by chopping of heading-stage silage. The steers fed chopped, but not those fed long dough-stage silage, selected for starch in the WCBS (P < 0.001). The neutral detergent fibre (NDF) intake was 5% higher (P < 0.01) for heading than for dough-stage silage and was associated with lower concentration of indigestible NDF (96 v. 170 g/kg DM). Rate of intake of DM and NDF was 37% higher (P < 0.001), daily eating time was 24% shorter (P < 0.001) and daily chewing time was 8% shorter (P < 0.05) for chopped silage compared with long silage but there was no effect of maturity. Daily rumination time was not affected by treatments, whereas rumination and chewing times per kg NDF intake were 15% and 13% higher (P < 0.05), respectively, for dough-stage than for heading-stage silage. The proportion of faecal particles retained on a 1 mm screen was 30% higher (P < 0.001) due to chopping and 45% higher (P < 0.001) due to delayed harvest. Chopping the dough-stage silage reduced the proportion of grain in faeces from 97 to 43 g/kg DM (P < 0.05) indicating higher starch digestibility. In conclusion, chopping increased DM intake of WCBS when harvested at dough stage but not at heading stage of maturity.

In goats, αs1-casein polymorphism is related to different rates of protein synthesis. Two genetic variants, A and F, have been identified as strong and weak alleles based on a production of 3·5 and 0·45 g/l of αs1-casein per allele. The aim of the trial was to test whether goats can select their diet as a function of their genetic aptitude to produce milk at different casein levels and whether this selection can influence milk production or composition. Two groups of 8 animals, homozygous for strong (AA) or weak (FF) alleles were housed in individual pens. Using a manger subdivided into five separate containers, the goats were offered daily for 3 weeks: 1·5 kg of alfalfa pelleted hay, 0·7 kg of whole barley, 0·7 kg of whole maize, 0·7 kg of whole faba bean and 0·7 kg of pelleted sunflower cake. Total dry matter intake was similar between groups and resulted in nutrient inputs much higher than requirements. On average, goats selected 86% of maize plus barley and only 46% of faba bean plus sunflower. Indeed, AA goats selected less faba bean compared with FF goats (37·2 v. 56·7% of the available amount; P=0·01); during week 2 and week 3 they significantly increased maize selection (respectively for week 2 and week 3: 94·9 and 99·1% v. 85·3 and 87·3%) thus increasing the ratio between the high-energy feeds and the high-protein feeds (2·41 v. 1·81, P=0·023). As for true protein, the high soluble fraction (B1) and the indigestible fraction (C) were lower in the diet selected by AA goats (respectively in AA and FF groups: B1, 7·85 v. 9·23% CP, P<0·01; C, 6·07 v. 6·30% CP, P<0·001); these diet characteristics can be associated with lower losses of protein. Milk production, being similar in AA and FF groups when goats were fed with a mixed diet, significantly increased in AA group, when free-choice feeding was given (mean productions: 1198 v. 800 g/d, P<0·01). Casein content was higher in AA group than in FF group (2·70 v. 2·40%, P<0·01) whereas milk urea was higher in FF group (59·7 v. 48·8 mg/dl, P<0·01). In conclusion, when the animals were free to select their diet, their higher genetic aptitude to produce casein seemed to adjust their energy and protein dietary input in qualitative terms, thus leading to an increase in milk production and a decrease in milk urea. These results seem to demonstrate that interactions probably occurred between genetic polymorphism at the αs1-casein locus, diet selection and the efficiency of nutrient transformation into milk.

Voluntary food intake and the selection between foods are important subjects especially in ruminants in view of the economic importance of this class of animal and the complex digestive system with its attendant metabolic peculiarities. There is evidence that intake is limited by the capacity of the rumen as well as by metabolic factors; some theories assume that intake is controlled by the first limiting factor but this is not satisfying on physiological grounds and there is evidence that signals from feedback factors are integrated in an additive manner. It is now well established from research in which animals are given the chance to learn the metabolic consequences of eating food with a particular sensory profile, including a choice of foods, that animals including ruminants can adjust their diet, both quantitatively and qualitatively, to their nutrient requirements. It is proposed that they do this in order to minimise the total of the discomfort generated by the several signals from various body systems. The learning process is aided by the considerable day-to-day variation often seen in the intake of individual animals. An optimisation model is proposed and presented in a simple form, involving the addition of discomforts (calculated as the square of the deviation of the supply of metabolisable energy, crude protein and neutral-detergent fibre) and iterative elucidation of the intake at which total discomfort is minimal. With parameters appropriate for growing lambs the model provides reasonable agreement with observations, both in terms of daily intake and selection between foods of different protein contents. Manipulation of food composition and of nutrient requirements produces predictions broadly in agreement with reality except that protein deficiency has less severe consequences for the model than for real animals; it is proposed that protein deficiency be given more weighting than protein excess, and this may be true for other resources as well. This model is proposed as a philosophy and a starting point for further development and is not purveyed as a complete, working model. It nevertheless provides support for the concept of total minimal discomfort as a suitable base from which to view the control of intake and selection in all animals.

Large herbivores play a major role in shaping vegetation community dynamics through selective consumption of particular plants and plant communities. An understanding of the factors influencing diet selection at the level of individual bites (‘bite scale’) is important for prediction of the impact of herbivores on vegetation at the habitat scale. Bite-scale diet selection represents an integration of the twin goals of maximizing nutrient intake and minimizing toxin intake. Recent research with ruminants in pen-fed situations has shown that animals are able to make choices between artificial foods that maximize growth and other production variables. The role of post-ingestive feedback as an important mechanism for allowing animals to assess the nutritional quality of particular foods, and so select optimal diets, has been recognized in a number of recent experiments. Our understanding of the role of toxin intake minimization in diet selection decisions is more rudimentary. An important advance in the last decade has been the acknowledgement of the role of post-ingestive feedback and learning as a mechanism for avoidance of dietary toxicity. Further research is required to assess the importance of these processes in relation to free-grazing animals. The extent to which an understanding of bite-scale diet selection can be used to predict habitat utilization is not well understood. At the habitat scale additional factors such as predator avoidance, social constraints, avoidance of parasitism and microclimatic effects have an important influence on foraging decisions. Future research needs to focus on developing a quantitative understanding of such decisions at the habitat scale.

Conditioned food aversion is a powerful experimental tool to modify animal diets. We have also investigated it as a potential management tool to prevent livestock from grazing poisonous plants such as tall larkspur (Delphinium barbeyi), white locoweed (Oxytropis sericea) and ponderosa pine (Pinus ponderosa) on western US rangelands. The following principles pertain to increasing the strength and longevity of aversions: mature animals retain aversions better than young animals; novelty of the plant is important, although aversions can be created to familiar plants; LiCl is the most effective emetic, and the optimum dose for cattle is 200 mg/kg body weight; averted animals should be grazed separately from non-averted animals to avoid the influence of social facilitation which can rapidly extinguish aversions. Social facilitation is the most important factor preventing widespread application of aversive conditioning. When averted animals see other animals eat the target food they will sample it, and if there is no adverse reaction they will continue eating and extinguish the aversion. However, if averted animals can be grazed separately, aversions will persist. Aversive conditioning may provide an effective management tool to prevent animals from eating palatable poisonous plants that cause major economic loss.

In the present paper we deal with the problems of explaining and predicting diet selection of animals under controlled conditions, i.e. conditions that can be described and in which any influences of the environment can be either controlled or at least monitored. Diet selection is considered within an integrative framework of feeding behaviour that views both food intake and diet selection as an outcome of the animal's internal state and knowledge of the feeding environment. Three questions that arise from the framework are considered: (1) how do animals learn about foods available to them as a choice? (2) what changes in internal state affect diet selection? (3) how much time is needed for a change in the animal's internal state to be detected and for it to react to this change through a modification of its diet selection? It is proposed that animals have developed behavioural mechanisms that allow them to recognize foods on the basis of their nutritional as well as other properties. The rate at which animals learn about foods depends largely on the extent of the animal's deficiency and on the extent of the post-ingestive consequences induced by the foods. There is little evidence that animals modify their diet selection in response to short-term systemic fluctuation of their internal environment. On the other hand, long-term changes in the internal state of the animal lead to consequent long-term changes in diet selection. The time needed for a change in diet selection to be observed depends on the deviation created in the animal's internal state, either as a result of a physiological change or as a consequence of feeding. Thus, a more appropriate question to consider is not 'what time period matters to the animal?' but 'how much change or deviation in the internal state is the animal prepared to accept?'

The ability of laying hens to adjust their intake of available P (AP) was investigated with a maize–soyabean diet fed to forty-eight individually caged birds in a 2×4 factorial experiment. From 19 to 25 weeks of age (phase 1) twenty-four birds were fed a normal-P (NP) diet (2·2 g AP/kg DM) and twenty-four were fed a low-P (LP) diet (1·1 g AP/kg). LP eggs were lighter (51 v. 54 (sem 1·0) g; P>0·05), providing evidence that the LP diet was deficient in AP. From 25 to 28 weeks of age six hens from each phase 1 treatment were fed either the NP or LP diet alone or a choice of the LP and NP feeds or a choice of the LP feed and a phytase-supplemented (PP) feed (LP diet with 400 microbial phytase units/kg). With a choice of the NP and LP feeds, the hens fed the LP diet in phase 1 ate a smaller proportion of the LP feed (34 (sem 12·0)%) than the hens fed the NP diet in phase 1 (72 (sem 12·0)%; P>0·05), showing that P deficiency influenced subsequent selection for AP, i.e. an appetite for P was demonstrated. In those birds offered the LP and PP feeds, the presence of phytase in one of the two feeds significantly alleviated the effect of P deficiency on egg and body weights. The proportion of the LP diet chosen was not significantly affected by phase 1 treatment; it was not necessary for the hens to eat more than 50% of PP feed.

The short-term consumption of foods that differed in energy density (ED) and/or NaHCO3 supplementation, on subsequent food intake and diet selection in sheep were measured. Thirty sheep weighing 35·9 (SD 2·89) KG WERE USED. TWO FOODS WERE FORMULATED: H HAD 11 AND L HAD 8 MJ METABOLIZABLE ENERGY/KG FRESH MATTER. FOUR FURTHER FOODS WERE FORMULATED BY ADDING EITHER 40 G NAHCO3/KG OR 16·5 G NACL/KG TO FOODS H AND L. NACL WAS ADDED TO GIVE THE SAME NA CONCENTRATION AS WITH 40 G NAHCO3/KG TO CONTROL FOR ANY EFFECTS OF NA PER SE. IN A PRELIMINARY TEST, IT WAS FOUND THAT A 2 H CONSUMPTION OF FOOD H SUPPLEMENTED WITH NAHCO3 COULD BUFFER POTENTIAL IMPACT ON THE RUMEN ENVIRONMENT OF SUBSEQUENT CONSUMPTION OF FOOD H ALONE (AS JUDGED BY RUMEN PH AND ACID-BUFFERING CAPACITY); HOWEVER, IT WAS NOT AS EFFECTIVE AS THE CONSUMPTION OF FOOD L ALONE IN DOING SO. EACH FOOD TREATMENT WAS OFFERED TO ONE OF SIX GROUPS (N 5) FOR 2 H FOLLOWING 16 H OF FOOD DEPRIVATION. SHEEP WERE THEN OFFERED A CHOICE BETWEEN H AND L FOR A FURTHER 6 H. SUPPLEMENTING H OR L WITH EITHER NAHCO3 OR NACL HAD NO SIGNIFICANT EFFECT ON EITHER INTAKE OR DIET SELECTION. ED SIGNIFICANTLY (P<0·01) AFFECTED INTAKE DURING THE 2 H SINGLE FEEDING PERIOD, WITH SHEEP OFFERED H OR L CONSUMING 540 AND 663 (sed 37) g respectively, but had no effect on subsequent intake during the choice period. During the choice period all sheep showed a preference for food H, but sheep previously offered L selected significantly more H (0·873 g/g) than sheep previously offered H (0·544 (sed 0·028) g/g; P<0·001). It is concluded that short-term consumption of foods that differ in ED, and hence in their potential impact on the rumen environment, significantly affects subsequent diet selection. This is in agreement with the hypothesis that ruminant animals select a diet to help maintain the rumen environment within a certain physiological range. Food H with 40 g NaHCO3 added/kg may not have been sufficient to affect subsequent diet selection. It is suggested that larger, rather than smaller, changes in the rumen environment achieved through previous feeding should be expected to alter subsequent diet selection.

Rumen microbial degradation is an important route for detoxification of secondary plant compounds encountered in the diets of free-grazing ruminants. Exposure to diets containing particular secondary plant compounds can lead to increased rates of secondary compound degradation in the rumen. An experiment was conducted to determine whether rumen adaptation to oxalic acid would influence the diet selection of goats offered choices between plant species differing in their oxalic acid content. Twelve adult female goats were divided into two groups of six animals each. One group received a daily oral dose, in gelatin capsules, of 0·6 mmol oxalic acid/kg live weight per d throughout the experiment while the other group received placebos consisting of empty gelatin capsules. After an adaptation period of 8 d, the animals were allowed to graze a mixture of spinach (rich in oxalic acid) and cabbage (low in oxalic acid) for 7 h/d on two consecutive days per week during four consecutive 1-week periods. Intervening days were spent on grass pasture. Diet composition and intake were measured using cuticular wax n−alkanes as internal markers. Results showed that adapted goats included a higher proportion of spinach in their diet (P < 0·05) although absolute intakes of spinach were the same for the two groups. Goats in the oxalic-acid-adapted group consumed less cabbage than control animals (P < 0·05) suggesting that adaptation to oxalic acid at the rumen level may have interfered with detoxification of cabbage-derived secondary plant compounds. Voluntary intake increased progressively through the four experimental periods (P < 0·001) with a tendency for higher intakes among control than among adapted animals (P < 0·1). The experiment demonstrates how differences in the rate of degradation of secondary plant compounds may influence diet selection in ruminants.