Search

We determined if supplemental macronutrients or activated charcoal influenced intake of sagebrush, a shrub that contains terpenes. Sheep (lambs 3 mo age, 33 kg) and goats (kids 5 mo age, 28 kg) were fed supplements high in energy, high in protein, or offered a choice between the two supplements before and after receiving sagebrush for 4 h/d. The effect of activated charcoal, a compound that reduces bioavailability of toxins, was assessed by offering charcoal mixed with a carrier or the carrier alone while animals consumed sagebrush for 4 h/d. The effects of supplemental macronutirents were tested before and after charcoal supplementation. Activated charcoal did not increase intake of sagebrush by sheep or goats (P > 0.05), but supplemental macronutrients influenced intake of sagebrush by both species (P < 0.05). Intake of sage-brush was markedly higher for animals fed the high-protein supplement than for animals fed the high-energy supplement (P < 0.05). Lambs (before and after charcoal supplementation) and kids (after charcoal supplementation) offered a choice between the two supplements used sagebrush at levels comparable to animals supplemented with protein. They consumed more CP than animals fed the high-energy supplement and more ME than lambs fed the high-protein supplement (P < 0.05). Thus, providing a choice between supplements enhanced intake of sagebrush and enabled animals to achieve a more balanced intake of macronutrients. Preference for supplement and sagebrush also depended on an animal's immediate past history with a supplement. Sheep and goats previously fed the high-energy or the high-protein supplement preferred, respectively, the high-protein or the high-energy supplement when given a choice (P < 0.05). The changes in preference for supplement caused a decreased intake of sagebrush by groups that preferred the energy concentrate, and an increased intake of sagebrush by groups that preferred the protein concentrate. Thus, supplements of different energy and protein densities consistently altered intake of sagebrush by sheep and goats.

Tannins are a heterogeneous group of phenolic polymers that can induce detrimental effects when consumed by herbivores. Polyethylene glycol (PEG) binds to tannins and thus attenuates their negative effects. Our objective was to determine whether sheep actively seek PEG when fed tannins and thus modify their foraging location as a function of the spatial distribution of PEG. Lambs were first trained to recognize the beneficial effects of PEG by offering a meal high in quebracho tannin (QT), which presumably caused malaise, and then PEG (MW, 3,350), which presumably led to recovery from malaise. Animals were then tested in an experimental area where they could forage at two different locations that contained in Trial 1 1) PEG and QT in adjacent food boxes (PEG+QT) or 2) QT and in Trial 2 1) PEG or 2) QT. Preference for foraging locations was tested under three conditions: 1) no preload meal, 2) a meal high in quebracho tannin 1 h before testing, and 3) a basal diet high in quebracho tannin. Lambs spent more time (P = 0.012) and ate more tannin-containing food (P = 0.022) at locations where PEG was present (PEG+QT) than where it was absent (QT; Trial 1). Lambs responded to increased tannins in their bodies (Conditions 2 and 3) by reducing intake (P = 0.0001; Trial 2) at sites containing only QT and by increasing intake (P = 0.0001, Trial 1; P = 0.001, Trial 2) and foraging time (P = 0.072, Trial 1; P = 0.0001, Trial 2) at locations where PEG was available. Thus, PEG influenced foraging location by sheep consuming quebracho tannin. Accordingly, it may be possible to formulate range blocks that enable herbivores to ingest PEG in tannin-rich habitats. Strategic distribution of PEG in those habitats may attract animals to underutilized feeding sites.

We determined how a food's flavor and digestibility, along with an animal's recent experiences, influenced food preference and intake. In three experiments, pregnant heifers were fed a basal ration (7.75 kg/animal) of alfalfa, barley, corn silage, and a vitamin/mineral supplement from 1500 to 2200. Exp. 1 determined the influence of recent exposure to flavored straw. Animals were divided into two groups (n = 16/group) and fed either untreated or ammoniated straw with digestibilities of 43 and 58%, respectively. Within each group, half of the heifers were fed maple-flavored straw and the other half were fed coconut-flavored straw from 1100 one day to 0900 the next day, with no base ration. We then offered straw in both flavors from 1000 to 1200 for the next 5 d. Animals fed maple-flavored straw for 1 d generally preferred coconut- over maple-flavored straw for the next 5 d, whereas animals previously fed coconut-flavored straw preferred maple-flavored straw (P < 0.001). The change in preference was stronger when animals were fed untreated compared with ammoniated straw. Experiments 2 and 3 determined the influence of offering straw in different flavors, either in sequence (Exp. 2) or simultaneously (Exp. 3). In Exp. 2, we offered heifers (n = 16) straw in three flavors (maple from 0900 to 1100, coconut from 1100 to 1300, and unflavored from 1300 to 1500) and compared their intake with that of heifers (n = 16) offered unflavored straw throughout the day. In Exp. 3, we compared intake of heifers (n = 16) simultaneously offered straw in three flavors (coconut, maple, and unflavored) with that of heifers (n = 16) offered only unflavored straw from 1000 to 1500. In both experiments, straw intake and preference differed between heifers offered straw in a variety of flavors as opposed to only unflavored straw (P < 0.05), but animals fed a variety of flavors did not consistently eat more than those fed only one flavor. During a post-trial preference test, heifers previously restricted to straw in one flavor for 5 d preferred straw in alternative flavors, whereas heifers fed straw in all three flavors preferred unflavored straw. Changes in preference were stronger for heifers fed untreated compared with ammoniated straw. Collectively, our results suggest that palatability, as evidenced by changes in preference and intake, is dynamic and depends on a food's flavor and nutritional quality and an animal's recent experiences with the food.

Research findings and management recommendations typically emphasize responses of the "average" individual, yet more than half of the animals in a group may differ significantly from the mean regarding food preference and intake. The productivity of a herd may be adversely affected if animals differing from the mean are fed a uniform diet formulated to meet the needs of the "average" individual. We compared the intake and performance of beef calves offered a choice or no choice among foods. Diets consisted of ad libitum access to either a chopped, mixed ration of rolled barley (31.3%), rolled corn (31.3%), corn silage (15.5%), and alfalfa hay (18.9%) (n = 16 calves) or a choice among those foods offered individually (n = 15 calves). Averaged across the 63-d trial, the two groups did not differ in ratios of protein to energy ingested (43 vs 43 g CP/Mcal ME; P = 0.50), but preference for foods high in energy or protein varied markedly for animals fed free-choice: on d 21 they had protein:energy ratios higher than those of animals fed the mixed ration, on d 2 the ratios were equal, and on d 40 they had protein:energy ratios lower than those of animals fed the mixed ration. Throughout the trial, no two animals consistently chose the same ingredients, and none selected a diet similar to the nutritionally balanced mixed ration, yet each animal ate a diet adequate to meet its needs. Animals offered the mixed ration tended to eat more than animals offered a choice (109 vs 102 g/kg MBW/d; P = 0.10), but they did not gain at a faster rate (0.89 vs 0.92 kg/d; P = 0.65). Gain/unit of food consumed also was similar for both groups (0.09 vs 0.10 kg/kg; P = 0.38). However, food cost/day was higher for animals fed the mixed ration than for those offered a choice ($1.58 vs $1.36; P = 0.03). Consequently, cost/kilograms of gain was higher for the mixed ration than for the choice group ($1.84 vs $1.49/kg; P = 0.045). These findings suggest that 1) animals can more efficiently meet their individual needs for macronutrients when offered a choice among dietary ingredients than when constrained to a single diet, even if it is nutritionally balanced; 2) transient food aversions compound the inefficiency of a single mixed diet by depressing intake even among animals suited to that nutritional profile; and 3) alternative feeding practices may allow producers to efficiently capitalize on the agency of animals, thus reducing illness and improving performance.

Tannins decrease food intake by reducing digestion and by causing illness, whereas polyethylene glycol (PEG) attenuates the aversive effects of tannins. Our objective was to determine whether sheep recognize the benefits of ingesting substances such as PEG when consuming tannins. If so, then ingestion of PEG should be 1) PEG-specific, 2) a function of previous experience with recovery from tannin-toxicosis, and 3) dependent on the presence/absence of tannins. During conditioning, lambs in Group 1 (n = 10) were offered a meal of high-tannin food, which presumably caused malaise, and then offered PEG (molecular weight, 3,350), which presumably led to recovery from malaise. Subsequently, lambs ingested a control food (wheat straw) that did not have the "medicinal" effects of PEG in the absence of the tannin diet. In contrast, lambs in Group 2 (n = 10) ingested PEG in the absence of the tannin diet, and they ingested the tannin diet only in association with wheat straw. Ingestion of PEG and straw by both groups of lambs increased as a function of the presence of tannins in the diet (P < 0.05). However, when offered a choice among the tannin diet, PEG and straw, or when given the tannin diet and then offered a choice between PEG and straw, lambs trained to associate PEG with tannins ate more PEG than lambs that ingested PEG without tannins (P < 0.05). The responses were apparently PEG-specific; straw intake did not differ between groups of lambs during testing (P > 0.05), and differences in PEG intake disappeared in the absence of tannins (P > 0.05). In summary, our results suggest that lambs fed high-tannin diets discriminated the effects of PEG from those provided by a "nonmedicinal" food (straw). Thus, it may be possible to formulate PEG supplements that allow herbivores to self-regulate intake of PEG under extensive management conditions.

Ruminants select foods higher in nutrients and lower in toxins than the average of what is available whether fed in confinement or foraging on pastures or wild lands. Our objective was to better understand how sheep learn to select the most nutritious foods when they concurrently ingest different foods. We hypothesized that novelty and generalization are two ways sheep discriminate among foods. The first experiment determined whether lambs offered two foods in a meal (one a novel-flavored food and the other a familiar food) acquired a preference for the novel-flavored food following an infusion of energy (starch) into the rumen. Lambs did not increase preference for the novel-flavored food when the amount of starch infused (150 g) was held constant (P > 0.05). However, when the amount of starch infused was made directly proportional to the amount of novel-flavored food ingested, lambs quickly formed a preference for the novel-flavored food (P < 0.001). The second experiment determined whether lambs generalized preferences from familiar to novel foods. Lambs fed coconut-flavored milo grain for 51 d subsequently preferred coconut-flavored straw to plain straw (P < 0.001). Conversely, lambs that were not fed coconut-flavored milo grain preferred plain to coconut-flavored straw (P < 0.001). Lambs infused with a source of energy (starch, 100 g/[lamb x d]) consumed more coconut-flavored straw and less plain straw than lambs that had no infusions or lambs infused with a toxin (LiCl, 100 mg/kg BW; P < 0.05). Lambs infused once with LiCl had the lowest preference for coconut-flavored straw (P < 0.05). Lambs conditioned with starch had the highest intakes of coconut-flavored straw, even after starch infusions were suspended (P < 0.1). When straw was replaced by an energy-dense food (wheat grain), all lambs equally preferred coconut-flavored grain to plain grain (P < 0.001). Collectively, our results suggest that lambs learned to prefer a novel-flavored food when the amount of energy reward was contingent on the amount of novel-flavored food ingested, and that they generalized from familiar to unfamiliar foods based on common flavor cues. Once generalization occurred, postingestive feedback from nutrients and toxins calibrated preference according to the food's utility.

Ingestion of condensed tannins decreases feed intake in ruminants. Polyethylene glycol (PEG) forms high-affinity complexes with tannins. In two experiments carried out on Holstein heifers, quebracho (Q) from the Aspidosperma quebracho served as source of condensed tannins. The aims of the study were (i) to quantify the effect of Q on feed intake and eating behaviour in cattle fed complete mixed diets (CMDs); (ii) to clarify if changes induced in ingestive behaviour and feed intake by Q in cattle can be reversed by feeding PEG; and (iii) to clarify if the decrease in feed intake is associated with short-term (astringency, post-ingestive malaise) or longer-term effects. In experiment 1, 500 g/day of Q was found to be the minimal dose that decreased feed intake in heifers. A ratio of PEG:Q equal to 1:12.5 did not fully restore feed intake. In experiment 2, four heifers received a random sequence of four rations in a Latin-square design with feeding cycles of ca. 7 days: CMD containing no supplements (C), or supplemented with 625 g/day of Q without PEG (Q), with 625 g/day of Q and 250 g/day of PEG (Q-PEG), or with 250 g/day of PEG without Q (PEG). Individual rations were continuously weighed in the trough and the behaviour of heifers was observed for 180 min after distribution of CMD. Overall, feeding Q was associated with lowered feed intake and shorter duration of eating bouts, mainly of the first eating bout, immediately after distribution of the diet. A larger portion of the diet was consumed subsequent to 180 min after distribution in Q-fed heifers. Eating rate and the water to food ratio were not affected by Q. The effects of Q on feed intake were attenuated by feeding PEG. Heifers adapted effectively to condensed tannins by increasing the number of eating bouts and the portion of diet consumed subsequent to 180 min after distribution, so that no differences in feed intake were noted on the last day of each feeding cycle. Data are interpreted to show that: (i) negative effects of Q on feed intake derive from astringency of CT and short-term post-ingestive malaise; (ii) the increased number of eating bouts and their wider partition throughout the day are means to preserve the ruminal environment in Q-fed heifers; (iii) PEG has the potential to neutralize negative effects of condensed tannins in cattle.

Ruminants eat a variety of foods from different locations in the environment. While water, cover, social interactions, and predators are all likely to influence choice of foraging location, differences in macronutrient content among forages may also cause ruminants to forage in different locations even during a meal. We hypothesized that lambs forage at locations containing foods that complement their basal diet and meet their nutritional needs. Based on this hypothesis, we predicted that lambs (n=12) fed a basal diet low in protein and high in energy would forage where a high-protein food (Food P) was located, and that lambs (n=12) fed a basal diet low in energy and high in protein would forage where a high-energy food (Food E) was located. Food P was a ground mixture of blood meal (50%), grape pomace (30%), and alfalfa (20%) that contained 47% crude protein (CP) and 2.211 Mcal/kg digestible energy (DE). Food E was a ground mixture of cornstarch (50%), grape pomace (30%), and rolled barley (20%) that contained 6% CP and 3.07 Mcal/kg DE. Food P provided 212 g CP/Mcal DE, whereas Food E provided 20 g CP/Mcal DE. Lambs growing at a moderate rate require 179 g CP and 3.95 Mcal DE. During Trial 1, we determined if lambs foraged to correct a nutrient imbalance, and if they preferred a variety of foods (Foods P and E) to only one food at a location (Food P or E). During Trial 2, we determined if nutrient-imbalanced lambs foraged in the location with the food that corrected the imbalance when the location of the foods changed daily. During Trial 3, lambs were offered familiar foods (Foods P and E) at the location furthest - and novel foods (wheat and soybean meal) at the location nearest - the shelter of their pen. During all three trials, lambs foraged most at the location with the food that contained the highest concentration of the macronutrient lacking in their basal diet, but they always ate some of both foods. Lambs did not feed exclusively at the location with a variety of foods (P and E). Rather, they fed at the location nearest the shelter that contained the macronutrient lacking in their diet. As availability of the food with the needed macronutrient declined in one location, lambs moved to the nearest location that had food with the needed macronutrient. When food that complemented their basal diet was moved to a different location, lambs foraged in the new location. Collectively, these results show that lambs challenged by imbalances in energy or protein selected foods and foraging locations that complemented the nutrient content of their macronutrient imbalanced basal diets.

Supplemental polyethylene glycol (PEG) increases intake when animals eat foods high in tannins, but little is known about how PEG affects preference for foods that vary in concentrations of macronutrients and tannin. We investigated how varying macronutrients and tannins (commercially available extracts from quebracho trees) affected food intake, and we assessed the degree to which PEG (MW 3350) affected intake of tannin-rich foods by sheep. From 0715 to 1800 daily, lambs were offered diets that varied in concentrations of macronutrients: high energy/low protein (75% barley/25% alfalfa), medium energy/medium protein (35% barley/65% alfalfa), and low energy/high protein (100% alfalfa). Preference for these diets was determined in the absence of tannin, and then, in Trials 1 to 3, tannin was added in increasing concentrations (from 5 to 20%) to the diets with high and medium levels of energy. In Trial 4, tannin (10%) also was added to the low-energy diet. Lambs were supplemented with either 50 g of PEG mixed with 50 g of ground barley or 50 g of ground barley alone from 0700 to 1715 daily; lambs always consumed all of these supplements. In the absence of added tannins, all lambs preferred high energy/low protein > medium energy/medium protein > low energy/high protein. As tannin levels increased, preference for the high- and medium-energy foods decreased, and all lambs preferred foods that were lower in tannins and higher in protein. Lambs supplemented with PEG ate more macronutrients and tannins than unsupplemented lambs, and the effect became increasingly apparent as tannin levels increased from Trials 1 to 4. We conclude that the effectiveness of supplemental PEG may be low if alternative forages are equal or superior in nutritional quality and contain fewer metabolites with adverse effects. In such cases, animals would likely prefer alternatives to high-tannin foods.

Tannins occur in many plant species, and they often suppress intake by reducing nutrient availability or by causing malaise. Polyethylene glycol (PEG) binds to tannins and may thereby increase the availability of macronutrients and decrease malaise. Supplemental PEG increases intake of tannin-containing plants by sheep, goats, and cattle. Given the strong response to supplemental PEG, we speculated that animals might self-regulate their intake of PEG when offered foods high in tannins. The objective of the first experiment was to determine if the amount of supplemental PEG (0, 25, 50, 75, or 100 g; molecular weight, 3,350) affected intake by lambs of a food (milo-tannin mix) containing 20% quebracho tannin. There was a linear relationship (Y = 272 + 1.2X; R2 = .86; P = .023) between the amount of supplemental PEG ingested and the subsequent intake of milo-tannin food by lambs. The objective of the second experiment was to determine whether lambs self-regulated intake of PEG when fed a ration that contained 0, 5, 10, 15, or 20% quebracho tannin and whether they adjusted their intake of PEG when tannin was removed from the diet. There was a positive relationship between the amount of PEG ingested and intake of food and tannin (P = .0001). Lambs fed high-tannin diets ate more PEG than controls (P = .03). Lambs fed the 20% tannin diet ate the most PEG, and controls ate the least PEG. Tannin limited intake of the diets, but PEG attenuated the response to a great degree (P = .065). Immediately after tannin was removed from the ration, lambs that formerly had been fed the 20% tannin ration ate more PEG than lambs fed the other rations (P = .0075). Ten of the lambs (5 from the 20% tannin group, 1 from the 15% tannin, and 2 each from the 10 and 5% groups) continued to eat PEG for 7 d after tannin was removed from their ration. When they were tested again 6 wk after the trial and offered tannin-free diets, their intake of PEG had decreased.

We hypothesized that the development and extinction of preferences for flavored foods paired with infusions of starch depends on the delay between food ingestion and starch administration and on the frequency of starch administration. To test this hypothesis, we divided 24 lambs into three groups of 8 lambs/group. On even days, half of the lambs were offered chopped wheat straw flavored with a distinctive flavor, whereas the other half received straw with a different flavor. Starch (12% of the daily DE required by the lambs) was infused into the rumen immediately after lambs began to eat straw in Group 1 (150 g starch), immediately after lambs began to eat straw (75 g) and 30 min later (75 g) in Group 2, and when straw refusals were collected 1 h after lambs began to eat straw in Group 3 (150 g). On odd days, the flavors were switched and infusions of starch were replaced with water. After 8 d of conditioning, starch infusions were suspended, and lambs were offered a choice of wheat straw with the two distinctive flavors. All lambs acquired strong preferences for flavored straw paired with starch (P< .001), but the delay between straw ingestion and starch infusion affected preferences (P<.05). Lambs that received starch 1 h after beginning to eat straw (Group 3) had lower preferences than lambs that received starch immediately after beginning to eat straw (Groups 1 and 2). Lambs that received two starch infusions (Group 2) did not show higher preferences for flavored straw than lambs that received only one starch infusion immediately after beginning to eat (Group 1). Lambs in our study retained preferences for at least 17 d when eating straw was no longer reinforced with starch (P<.001). However, preferences were less persistent for lambs that received starch immediately after beginning to eat straw (Group 1) than for lambs that received starch in two pulses (Group 2) or lambs that received starch 1 h after beginning to eat straw (Group 3). Collectively, our results support the hypothesis that the delay between food ingestion and starch administration affects food preferences of lambs. These findings suggest that manipulating the chemical characteristics of foods, which affects the kinetics of forage degradation in the rumen, can increase food preferences in ruminants.

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 studies of behavior and nutrition, we typically determine nutritional needs and formulate diets for the average member of the herd, not for specific individuals within a herd. Nevertheless, variation among individuals could affect productivity of the group if the diet diverges too much from what individuals at the extremes prefer to eat. Thus, it is important to determine the degree to which individuals within a group vary in their food preferences when offered alternatives. Our first objective was to determine the degree to which lambs differed in preference for foods high in energy (barley) or protein (alfalfa) (Exp. 1). When we offered lambs barley and alfalfa for ad libitum consumption for 25 d, individuals varied in the amounts of barley (range: 221 to 991 g/d) and alfalfa (range: 51 to 558 g/d) they consumed (P < .0001). At one extreme, individuals preferred a diet of 6% alfalfa and 94% barley; at the other extreme, individuals preferred a diet of 70% alfalfa and 30% barley. Our second objective was to determine whether lambs from Exp. 1 compensated, when fed a basal diet that was lower in alfalfa than they preferred, by ingesting foods higher in alfalfa when offered a meal (Exp. 2). Lambs were ranked according to the percentage of alfalfa (range from 6 to 70%) and barley (range from 94 to 30%) they ate during Exp. 1 and then assigned alternately to two treatments: 1) basal diet with similar proportions of alfalfa and barley consumed ad libitum (preferred diet) or 2) basal diet with 10% less alfalfa than consumed ad libitum (low-alfalfa diet). We then conducted three trials in which lambs fed the different basal diets were offered a meal for 15 min/d for 2 d of two foods that differed in barley and alfalfa. During Trial 1, when we offered barley and alfalfa, lambs in both groups preferred barley (138 g) to alfalfa (46 g) (P < .05). During Trial 2, when the test foods (barley and alfalfa) were diluted with grape pomace (20%), lambs fed the preferred basal diet ate more barley (116 vs 64 g) and less alfalfa (48 vs 87 g) than lambs fed the low-alfalfa basal diet (P < .05). During Trial 3, when we offered a food high in barley (80% barley and 20% pomace) and a food high in alfalfa (70% alfalfa, 14% cornstarch, and 16% pomace), lambs fed the preferred basal diet ate more of the high-barley food (124 vs 73 g) and less of the high-alfalfa food (45 vs 98 g) than lambs fed the low-alfalfa basal diet (P < .05). Collectively, these results illustrate that lambs varied greatly in their preferences for foods that differ in energy (barley) and protein (alfalfa), and that when their preferred basal diet was altered, lambs compensated by ingesting food that complemented their basal diet during a daily meal. The addition of grape pomace in Trials 2 and 3 reduced the protein content of the high-barley and high-alfalfa foods such that the high-barley food was only marginally adequate to meet needs compared with the high-alfalfa food. Lambs fed the low-alfalfa basal diet compensated by eating more of the high-alfalfa food than lambs fed the preferred basal diet.

We hypothesized that lambs discriminate between postingestive effects of energy and protein and associate those effects with a food's flavor to modify food choices. Based on this hypothesis, we predicted that 1) lambs would acquire a preference for a poorly nutritious food (grape pomace) eaten during intraruminal infusions of energy (starch) or protein (casein) and that 2) shortly after an intraruminal infusion of energy or protein (preload), lambs would decrease their preferences for foods previously conditioned with starch or casein, respectively. Thirty lambs were allotted to three groups and conditioned as follows. On d 1, lambs in each group received grape pomace containing a different flavor and water was infused into their rumens as they ate the pomace. On d 2, the flavors were switched so each group received a new flavor and a suspension of starch (10% of the DE required per day) replaced the water infusion. On d 3, the flavors were switched again, and a suspension of casein (2.7 to 5.4% of the CP required per day) replaced the starch infusion. Conditioning was repeated during four consecutive trials. Lambs in Trial 1 had a basal diet of alfalfa pellets (e.g., free access from 1200 to 1700) and 400 g of rolled barley. Lambs in Trials 2, 3, and 4 received a restricted amount of alfalfa pellets (990 g/d) as their basal diet. After conditioning, all animals received an infusion of water, and, 30 min later, they were offered a choice of the three flavors previously paired with water, starch, or casein. On the ensuing days, the choice was repeated, but starch, casein, and barley replaced the water preload. The nutrient density of the infused preloads was increased during consecutive trials. Lambs preferred the flavors paired with starch > water > casein during Trial 1 (P < .05) and the flavors paired with starch > casein > water during Trials 2 (P < .05), 3 (P < .001), and 4 (P < .001). Preloads of casein decreased preferences for flavors previously paired with casein (P < .10 [Trial 2]; P < .001 [Trial 3], and increased preferences for flavors paired with starch (P < .05 [Trial 2]; P < .001 [Trial 3]). Preloads of energy (barley) had the opposite effect (P < .05 [Trial 3]). These results indicate that lambs discriminated between the postingestive effects of starch and casein and associated the effects with specific external cues (i.e., added flavors) to regulate macronutrient ingestion.

In two experiments on appetitive learning we conditioned lambs, Ovis aries, to particular concentrations of a flavour by mixing the flavour with an energy-rich food that complemented their energy-poor diet. The lambs were subsequently offered energy-rich food with five different concentrations of the flavour (the concentration to which they were conditioned, two higher concentrations, and two lower concentrations). At these tests, the lambs consistently preferred the weaker flavours. This finding stands in contrast to earlier results on generalization gradients. In a third experiment, similarly designed to the other two, we tested for effects of a strong flavour on the behaviour of lambs when they were offered a novel nutritious food. Half of the lambs were offered unadulterated wheat, and the others strongly flavoured wheat. We found that the flavour in itself was initially aversive. We propose that the lambs' avoidance of foods with strong flavours may be an expression of a rule of thumb of the type 'given a choice, avoid food with strong flavours'. Such a rule could be part of a risk-averse foraging strategy displayed by mammalian herbivores, and which could be of particular importance when they encounter unfamiliar foods. Copyright 1998 The Association for the Study of Animal Behaviour

We conducted experiments to determine whether preference for barley was affected when lambs ate various amounts of barley and whether lambs ate more barley when it contained lasalocid and sodium bicarbonate (NaHCO3), both of which attenuate acidosis. In Exp. 1, lambs were assigned to two treatments (six lambs/treatment). For 2 d, lambs in two treatments were offered either 400 or 1,200 g of rolled barley from 0600 to 0700 as a preload meal. A preference ratio [PR = barley ingested/(total amount of alfalfa + barley ingested)] was calculated based on lambs' intake when offered a choice of 200 g each of rolled barley and alfalfa pellets hourly from 0700 to 1100. After the preload meal, lambs in Treatment 1 (400 g preload) showed equal preference for barley (.52) and alfalfa (.48) for 4 h on d 1 (P > .05); their preference for barley was less after the meal of barley on d 1 (.52) than on d 2 (.72), but their preference for barley declined between h 3 (.81) and 4 (.55) of d 2 (P = .11). Lambs in Treatment 2 (1,200 g preload) showed a low preference for barley on d 1 (.29) and 2 (.19) (P < .001). In Exp. 2, lambs were assigned to four treatments (six lambs/treatment): 1) rolled barley + NaHCO3 (2%) + lasalocid (33 ppm); 2) rolled barley + NaHCO3 (2%); 3) rolled barley + lasalocid (33 ppm); or 4) rolled barley. Intake of barley by lambs offered NaHCO3 + lasalocid (Treatment 1) was greater (P = .07) than that by lambs offered NaHCO3 (Treatment 2), whereas intake by lambs offered lasalocid (Treatment 3) was similar (P > .05) to that by controls. We conclude that eating barley too frequently or in excess caused a decrease in lambs' preference for barley and that NaHCO3 and lasalocid attenuated the aversion.

We conducted experiments to determine whether lambs fed grain prefer foods and solutions containing sodium bicarbonate (NaHCO3) and lasalocid, compounds capable of attenuating acidosis. In Exp. 1, we determined whether lambs fed barley preferred flavored rabbit pellets (RP) containing NaHCO3 and lasalocid. Lambs in two groups (n = 10/group) were fed increasing amounts of barley on d 1 to 12 (300 to 1,100 g) and again on d 23 to 34 (300 to 1,350 g). After ingesting barley on d 1 to 12, lambs were fed ground RP containing lasalocid and NaHCO3 (i.e., medicated) and flavored with either 2% onion (group 1) or 2% oregano (group 2). During d 23 to 34, lambs were fed unmedicated RP containing NaCl and flavored with either 2% oregano (group 1) or 2% onion (group 2). During preference tests on d 35 to 40, lambs fed grain preferred RP with NaHCO3 to RP with NaCl (151 vs. 96 g; P < .01). In the Exp. 2, we determined whether wheat ingestion affected consumption of aqueous solutions containing NaHCO3. In trial 1, 28 lambs were assigned to four treatments: 1) low-wheat + 2% NaHCO3, 2) high-wheat + 2% NaHCO3, 3) low-wheat + water, and 4) high-wheat + water. For 12 d from 0800 to 0830, lambs in treatments 1 and 3 were fed 300 g of wheat and lambs in treatments 2 and 4 were fed up to 1,300 g of wheat; fluids (NaHCO3 and water) were then offered from 0930 to 1230 daily. Lambs drank more NaHCO3 on the high- than on the low-wheat diet (1,332 vs 890 g; P = .03); water consumption was similar for lambs on the high- and low-wheat diets (1,675 vs 1,700 g; P > .10). In trial 2, lambs in treatments 3 and 4 were offered a solution containing 1.4% NaCl. For 13 d from 0800 to 0830, lambs in treatments 1 and 3 were fed 500 g of wheat and lambs in treatments 2 and 4 were fed up to 1,700 g of wheat. Lambs had access to fluids from 0800 to 1200 daily. Lambs drank nearly twice as much NaHCO3 solution on the high- than on the low-wheat diet (1,066 vs 572 g), whereas they drank only 1.4 times more NaCl solution on the high- than on the low-wheat diet (888 vs. 634 g; P < .001). Fewer lambs showed signs of acidosis in treatment 2 than in treatment 4 in trials 1 (2 vs 9) and 2 (7 vs 17). Collectively, these results are consistent with the hypothesis that lambs fed grain prefer substances that attenuate acidosis.

We addressed two questions involving food preference. First, we determined how a food's flavor and nutritional characteristics affected preference. In three trials, we offered lambs isonitrogenous foods differing in energy (trial 1, 90% TDN; trial 2, 100% TDN; trial 3, 110% TDN); each food was offered in apple and maple flavors. We hypothesized that preference for apple- or maple-flavored food would decrease with increasing duration of exposure (1, 2, or 4 d), and we speculated that the change in preference would intensify when food contained inadequate or excessive levels of energy. After eating food in one flavor, lambs preferred the alternative flavor, even after only a 1-d exposure, and preference for the alternative flavor was greater when the food had inadequate or excessive energy (P < .05). The second experiment determined whether eating a food with rapidly or slowly digestible sources of energy in the morning affected lambs' food preferences in the evening. We speculated that lambs fed rapidly digestible food in the morning may prefer a slowly digestible food in the afternoon because slowly digestible food better maintains nutrient status throughout the night or because preference for the rapidly digestible food decreases after exposure in the morning. We offered lambs isonitrogenous and isocaloric foods, that differed in rates of digestion, in apple and maple flavors. Lambs fed rapidly digestible food in the morning preferred slowly digestible food in the alternative flavor in the evening. However, lambs fed slowly digestible food in either flavor in the morning preferred slowly digestible food in both flavors in the evening (P < .05). These results show that lambs' preferences change as a result of food ingestion, and the degree of change in preference depends on the nutritional characteristics of the food. These findings further suggest food intake might be increased by providing a variety of foods to livestock on rangelands, pastures, or in confinement.

We hypothesized that volatile fatty acids in rumen fluid are feedback signals that can condition food preferences or aversions in sheep. Three predictions were tested based on this hypothesis: 1) low doses of sodium propionate or sodium acetate condition preferences, but high doses condition aversions (Exp. 1 and 2); 2) preferences are not caused by osmotic load (Exp. 3 and 4); and 3) low doses of mixtures of acetate:propionate condition preferences (Exp. 4). In Exp. 1, 2, and 4, lambs were divided into four groups (10 lambs/group), and lambs in Exp. 3 were divided into two groups (five lambs/group). In all experiments, alfalfa pellets were the basal diet. On even days, half of the lambs were offered chopped wheat straw containing a distinctive flavor, whereas the other half received straw with a different flavor. During straw ingestion, different groups of lambs received intraruminal infusions of different concentrations (4, 8, or 12% of the daily DE received) of sodium propionate (Exp. 1), sodium acetate (Exp. 2), NaCl at osmotic loads equivalent to those when propionate supplied 4% of the daily DE received (Exp. 3), or different proportions of sodium acetate:sodium propionate (55:45 or 75:25% of the DE of the infusion [4% of the daily DE received]), or equimolar amounts of NaCl (Exp. 4). On odd days, the flavors were switched, and no infusions were administered. After 8 d of conditioning, lambs were offered a choice of wheat straw with the two distinctive flavors. Lambs preferred the flavor paired with the lowest doses of propionate (P = .07) and acetate (P = .08) but avoided the highest doses (P < .001). Excesses of VFA may condition aversions due to increases in rumen fluid osmolality and(or) excessive rates of supply of energy or sodium to the rumen. Lambs also preferred flavored straw associated with combinations of acetate and propionate (P < .001), especially at the highest concentration of propionate (P = .10). Lambs avoided NaCl in Exp. 3 (P < .001) and did not form preferences for NaCl in Exp. 4 (P > .05). Thus, osmolalities were not responsible for flavor preferences. In conclusion, our results support the hypothesis that food preferences and aversions reside along a continuum that depends on the amount of VFA infused.

We suggested that food preference depends on the interplay between flavour and post-ingestive effects, and we predicted that protein-restricted lambs would acquire preferences for foods paired with supplemental sources of N, including urea (Expts 1 and 2), casein (Expt 3), and gluten (Expt 4). In each experiment, twenty lambs, in two groups of ten, were conditioned as follows: on odd-numbered days, lambs in group 1 received wheat straw (Expts 1, 3, and 4) or ground barley (Expt 2) flavoured with a distinctive flavour, and lambs in group 2 received the same food but with a different flavour. On even-numbered days, flavours were switched and lambs received capsules containing different amounts of urea (ranging from 0.12 to 0.92 g N/d), casein (ranging from 0.23 to 0.69 g N/d), or gluten (ranging from 0.23 to 0.69 g N/d). After conditioning period of 8 d, lambs were given a two-choice test to determine preference for flavours paired with N. In Expts 1 and 2, lambs preferred the flavours conditioned with urea at lower doses (0.12 g N/d in Expt 1, 0.23 and 0.46 g N/d in Expt 2), but they avoided the flavour associated with urea at the highest dose (0.23 g N/d in Expt 1 and 0.92 g N/d in Expt 2). In Expts 3 and 4, lambs avoided the flavours associated with the lowest doses of casein or gluten (0.23 g N/d), but they preferred the flavours paired with casein or gluten at higher doses (0.46 and 0.69 g N/d). After conditioning, N administrations were suspended and lambs in Expts 3 and 4 were offered a choice of the two flavours at weekly intervals for 2 weeks (extinction); preferences persisted during extinction. Collectively, these results suggest that the post-ingestive effects of N in different forms and concentrations influenced the development of food preferences by lambs.

We hypothesized that feed preference depends on the interplay between flavour and postingestive effects, and we tested two predictions based on this hypothesis: (1) lambs acquire preferences for poorly nutritious feeds paired with starch; and (2) preferences persist when starch is no longer administered. Twenty lambs were randomly allocated to two groups and conditioned as follows: on odd-numbered days, lambs in group 1 received onion-flavoured wheat straw and lambs in group 2 received oregano-flavoured wheat straw. On even-numbered days, the flavours were switched and starch (2.5-9.4% of the digestible energy received/d) was infused into the rumen of all animals during straw consumption. Four periods of 8 d of conditioning were performed and on the 9th day of each period all animals were offered a choice between onion- and oregano-flavoured straw. After conditioning, starch administration was suspended and lambs were offered onion- and oregano-flavoured straw at weekly intervals for 8 weeks (extinction). Lambs strongly preferred the flavoured straw paired with starch, and this preference persisted during extinction. Thus, these results suggest that the postingestive effects of energy play an important role in the development of feed preferences of ruminants.

We investigated how food deprivation affected preference of lambs for foods that varied in concentrations of nutrients and a toxin. Thirty lambs (10 lambs/treatment) were fed different amounts of alfalfa pellets (high in protein, marginal in energy for growth) as a basal ration (20, 40, or 60 g/kg body weight). Every morning, prior to ingesting the basal ration of alfalfa pellets, each lamb was offerend three foods for 15 min. The foods contained different amounts of energy and a toxin, depending on the addition of barley (energy) and LiCl (toxin) to alfalfa. The proportions of barley and LiCl changed every five days during the 25-day study. The results showed: (1) all lambs preferred food that was high > intermediate > low in energy (barley) in the absence of LiCl, but all lambs decreased consumption of foods high in energy as LiCl concentrations increased; (2) the greater the level of food deprivation, the lower the consumption of foods containing LiCl, even if the foods provided high levels of energy; (3) lambs moderately food deprived or fed ad libitum ingested more LiCl than lambs that were highly deprived; and (4) lambs quickly (15 min/day) regulated intake of foods in response to changes in barley and LiCl concentrations. Thus, our results suggest that the interaction between nutritional status and toxicosis plays an important role in food preference of lambs. Our findings also suggest that toxic plants may kill herbivores that lack nutritious alternative foods not only because the animals are forced to be less discriminating, but also because they are more susceptible to toxins.

We hypothesized that volatile fatty acids are feedback signals that condition food preferences in ruminants, and we tested two predictions based on this hypothesis: 1) low doses of propionate condition preferences for low-quality foods (Exp. 1 and 2) preferences are not caused by osmotic load (Exp. 2). In Exp. 1, lambs were offered chopped wheat straw flavored with either oregano or onion on odd days, whereas on even days flavors were switched and lambs received capsules containing sodium propionate. During four 8-d conditioning periods, the amounts of propionate delivered ranged from .7 to 1.4% of the daily DE intake (Period 1) or were fixed at .7% (Period 2) and 1% of the daily DE intake (Periods 3 and 4). After each 8-d conditioning period, lambs were offered oregano- and onion-flavored straw. Conditioning was then suspended and lambs were offered onion- and oregano-flavored straw at weekly intervals for 1 mo (extinction). Lambs preferred the flavor paired with propionate during conditioning (P < .001) and extinction (P < .07). During Exp. 2, a different group of lambs was conditioned as in Exp. 1, but sodium chloride was delivered at osmotic loads equivalent to those when propionate supplied .7% and 1% of the daily DE intake. Lambs strongly avoided the flavor paired with sodium chloride (P < .001). Thus, lambs acquired preferences for straw conditioned with doses of propionate typically considered ineffective in the regulation of food intake, and osmolalities generated by propionate did not cause, but probably attenuated, food preferences.

Ruminants eat a variety of foods, varying in toxins and nutrients, but no hypotheses adequately explain this behavior. We offer an explanation, one which encompasses avoidance of toxins and acquisition of nutrients. A key concept in this hypothesis is aversion, the decrease in preference for food just eaten as a result of sensory input (taste, odor, texture, i.e., a food's flavor) and postingestive effects (of toxins and nutrients on chemo-, osmo-, and mechano-receptors) unique to each food. On the basis of this hypothesis, we predicted lambs would prefer familiar and novel foods that complemented the macronutrient composition of their basal diet. To assess the validity of this prediction, we fed 10 lambs in each of three treatments different levels of ground barley (high in energy) and alfalfa (high in protein) as a basal diet. We then offered them daily a meal of three ground foods differing in proportions of barley and alfalfa (familiar foods) or wheat and rabbit pellets (novel foods). We found that lambs fed a basal diet high in energy (barley) preferred food lower in energy and higher in protein (alfalfa); those fed a diet high in alfalfa preferred food high in barley (P < .01). In addition, the higher the barley or alfalfa content of the basal diet, the greater the acceptance of novel foods high in alfalfa (i.e., rabbit pellets) or grain (i.e., wheat), respectively (P < .01). All lambs preferred foods high in wheat to rabbit pellets or alfalfa (P < 0.01), evidently because wheat is high in energy and it differs in flavor from barley, which was eaten repeatedly as part of the basal diet. On the basis of these results, we contend that lambs preferred familiar and novel foods that complemented the flavors and macro-nutrient contents of their basal diet.

Our objective was to better understand the importance of flavor and nutrients in food preferences of lambs. Three foods differing in flavor and nutritional quality were created by grinding and mixing grape pomace, barley, alfalfa pellets, and soybean meal in different proportions. food 1 (2.21 Mcal/kg DE, 8.1% DP), food 2 (2.42 Mcal/kg DE, 11.0% DP), and food 3 (2.68 Mcal/kg DE, 13.8% DP). Intake of each food, offered singly and together, was assessed when foods 2 and 3 were flavored with 1% onion or 1% oregano. Lambs (n = 24) preferred food 3 > 2 > 1, regardless of flavor (P < .05), and they continued to prefer food 3 > 2 > 1, even when they received the toxin LiCl after eating one of the three foods (P < .05). When offered a choice, lambs always ate substantial amounts of all three foods, even though they might have been expected to eat food 3 exclusively. We hypothesize selection of a varied diet resulted from a decrease in preference for food just eaten as a result of sensory input (taste, odor, texture, i.e., a food's flavor) and postingestive feedback (effects of nutrients and toxins on chemo-, osmo-, and mechano-receptors) unique to each food. Thus, we submit that offering different foods of similar nutritional value, offering foods of different nutritional value, and offering the same food in different flavors are all means of enhancing food preference and intake.

Ruminants eat an array of plant species that vary in nutrients and toxins. This selection makes intuitive sense, but no theories adequately explain this diversity. Some maintain it reduces the likelihood of overingesting toxins, whereas others contend it meets nutritional needs. Nevertheless, herbivores seek variety even when toxins are not a concern and nutritional needs are met. I offer another explanation for this behavior, one which encompasses the avoidance of toxins and the acquisition of nutrients. A key concept in this theory is aversion, the decrease in preference for food just eaten as a result of sensory input (a food's taste, odor, texture, i.e., its flavor) and postingestive effects (effects of nutrients and toxins on chemo-, osmo-, and mechano-receptors) unique to each food. Aversions are pronounced when foods contain toxins or high levels of rapidly digestible nutrients; they also occur when foods are deficient in specific nutrients. Aversions occur even when animals eat nutritionally adequate foods because satiety (satisfied to the full) and surfeit (filled to nauseating excess) represent points along a continuum, and there is a fine line between satiety and aversion. Thus, eating any food is likely to cause a mild aversion, and eating a food too frequently or in excess is likely to cause a strong aversion. Aversions are involuntary and are not the result of conscious decisions by an animal. Aversions yield benefits (e.g., obtain a balanced diet, reduce ingestion of toxic foods, optimize foraging and rumination times, sample foods, maintain a diverse microflora in the rumen) that are often mistaken as the cause of varied diets. In this article, I discuss the subtle ways in which aversions diminish preference and cause animals to eat a variety of foods.

Three kinds of memory help herbivores track changes in the environment. The first is the collective memory of the species with genetic instructions that have been shaped by the environment through millennia. This includes skin and gut defense systems. Auditory and visual stimuli and sensations of pain impinge upon the skin defense system that evolved in response to predation. The taste of food and the sensations of nausea and satiety are an integral part of the gut defense system that evolved in response to toxins and nutrients in plants. The second kind of memory in social mammals is represented by the mother, a source of transgenerational knowledge, who increases efficiency and reduces risk of learning about foods and environments. The third kind of memory is acquired by individual experience. Postingestive feedback from nutrients and toxins enables individuals to experience the consequences of food ingestion and to adjust food preference and selection commensurate with a food's utility. The three memories interact, each linking the past to the present, and collectively shape the present and future of every individual. Thus, the dynamics of foraging involves appreciating the uniqueness of individuals and subgroups of animals, each with their own genetic and behavioral history, and recognizing that foraging behaviors may not be stable, optimal, or even predictable in the conventional sense.

Livestock and range managers would have a powerful tool to direct utilization and modify plant communities if animals could be conditioned to eat specific foods or plants. We attempted to condition preferences for a low-quality forage through nutrient loading. Sheep were fed licorice or orange-flavored straw pellets then were gavaged with glucose or propionate (.381 Mcal, which amounted to approximately 13% of the daily maintenance energy requirement) or water. Four groups of ewes (n = 4) were arranged in a cross-blocked design such that each group received a unique energy/flavor combination: 1) propionate+licorice, 2) propionate+orange, 3) glucose+licorice, or 4) glucose+orange. On alternate days, each group received the other flavor plus water to create an internal control. At the end of 8 d of conditioning, preference for the two flavors was measured by two-choice preference tests. A second trial was conducted for 4 d in which the energy level was doubled to .762 Mcal. Low energy levels of either glucose or propionate did not create significant preferences. Propionate at the low-level caused satiety but at the high level conditioned an aversion to both flavors. This high level of propionate apparently caused malaise that was then associated with the taste of the flavors. The high level of glucose conditioned a preference. The high glucose treatment increased rumen microbial mass, the nutrients of which would have been absorbed in the lower gastrointestinal tract, and may have indirectly provided the positive nutrient feedback required to form a preference. There was a flavor preference for orange that was independent of the energy supplements.

Ruminants learn to avoid many foods that contain toxins by associating the flavor of the foods with aversive postingestive feedback. We hypothesized the emetic system is a cause of aversive feedback, and three experiments were conducted to determine whether antiemetic drugs (diphenhydramine, metoclopramide, dexamethasone) would attenuate food aversions caused by the toxicant lithium chloride (LiCl). Lambs were assigned to one of four treatments: antiemetics plus LiCl (A + L), antiemetics alone (A), LiCl alone (L), or neither antiemetics nor LiCl (C). The LiCl was administered immediately after sheep ate oats, wheat, and milo in Exp. 1, 2, and 3, respectively. The antiemetics were given 1 h before and at the time LiCl was administered to sheep. Lambs that received antiemetics (A and A + L) consistently ate more grain than lambs that did not receive the drugs (C and L) (Exp. 1, P < .08; Exp. 2, P < .05; Exp. 3, P < .08), and there was no interaction between antiemetics and LiCl. Thus, the results of all three experiments were consistent with the hypothesis that antiemetic drugs attenuate food aversions caused by the toxicant LiCl because sheep receiving antiemetic drugs (Group A + L) ate more grain than sheep not receiving the drugs (Group L). In addition, we suggest aversive postingestive feedback limited intake of grain because sheep receiving antiemetic drugs (Group A) ate somewhat more grain than sheep not receiving the drugs (Group C).

Food avoidance can be directly motivated in the sense that an animal refrains from eating a food because of an aversion to it, or indirectly motivated in the sense that an animal ingests little of one item because it prefers another. We studied whether the reluctance of a lamb to eat a shrub (Cercocarpus montanus) its mother avoided resulted from a socially induced aversion to C. montanus or from a socially induced preference for an alternative shrub (Amelanchier alnifolia). Each lamb (n = 6) in the treatment group was exposed with its mother to either A. alnifolia or C. montanus for 5 min, followed by 5 min of exposure to the other food for 5 d. Mothers in the treatment group avoided C. montanus because its ingestion had previously been paired with lithium chloride, but they readily ate A. alnifolia. Each lamb in the control group (n = 6) was exposed with its mother to only A. alnifolia. Following weaning, lambs in both groups strongly preferred A. alnifolia to C. montanus when offered a choice between the two species (P < .05), but they readily ate C. montanus when only C. montanus was offered. Accordingly, the data support the hypothesis that the low consumption of C. montanus occurred as a result of a socially induced preference for A. alnifolia, not as a result of a socially induced aversion to C. montanus. If food avoidance had been directly mediated, lambs whose mother avoided C. montanus would have completely avoided the shrub during testing, as in the case when acquired food aversions are directly mediated by toxins.

We discovered that a food aversion could be conditioned in anesthetized sheep. Sheep were allowed to eat a familiar food (alfalfa-grain pellets) for 30 min, and 90 min later they were given either an intraruminal (IR) injection of water (C), an IR injection of LiCl (L), anesthesia followed by an IR injection of water (A), or anesthesia followed by an IR injection of LiCl (A+L). Induction of anesthesia was by an intravenous injection of pentobarbitone sodium, and maintenance of deep anesthesia was by halothane. Sheep were maintained in deep anesthesia for 2 h to ensure that the effects of LiCl on the acquisition of a food aversion, which occur within about 1 h, were completed before they awakened. When tested 5 days later, sheep that received LiCl (treatments L and A+L) consumed less alfalfa-grain pellets than sheep that did not receive LiCl (treatments C and A) (241 g vs. 306 g; p = 0.057). Intake of sheep that were anesthetized (treatments A and A+L) did not differ from that of sheep that were not anesthetized (treatments C and L) (295 g vs. 252 g; p = 0.183). Nor was there an interaction between LiCl and anesthesia (p = 0.423). Thus, we conclude that changes in preferences for foods caused by postingestive feedback occur automatically every time food is ingested (i.e., they are noncognitive), and the kind and amount of feedback is a function of the match between the food's chemical characteristics and its ability to meet the animal's current demands for nutrients.

To better understand some of the mechanisms that control selection of novel foods differing in postingestive consequences, we offered goats current season's (CSG) and older (OG) growth twigs from the shrub blackbrush (Coleogyne ramosissima). CSG is higher than OG in nitrogen (1.04% v. 0.74%) and it is more digestible in vitro in goat rumen fluid (48% v. 38%). Nevertheless, goats acquire a preference for OG because CSG contains much higher levels of a condensed tannin that causes a learned food aversion. When CSG and OG were offered to goat naive to blackbrush, the goats did not choose either OG or CSG exclusively, but when they finally (1) ate more CSG than OG within a meal (averages of 44 g and 16 g, respectively) and (2) ate enough CSG within the meal to acquire an aversion (average of 44 g), they ingested less CSG than OG from then onward. Accordingly, the change in food selection resulting from postingestive feedback was influenced by the amount of each food ingested within a meal. This was further shown when we varied the amounts of CSG and OG that goats ingested within a meal, and then gave them by gavage the toxin lithium chloride (LiCl). They subsequently ate less of the food eaten in the greatest amount, regardless of whether it was CSG or OG. The salience of the flavor (i.e., taste and odor) of CSG and OG also played a role in the acquired aversion to CSG. Salience evidently was due to a flavor common to both OG and CSG that was more concentrated in CSG. We conclude that the relative amounts of different foods ingested within a meal, and the salience of the flavors of those foods, are both important variables that cause goats to distinguish between novel foods that differ in postingestive consequences.

If an animal experiences gastrointestinal malaise after eating a novel feed, it develops a dislike for the feed called a conditioned flavor aversion (CFA). Understanding flavor aversions is important for diet-training procedures, understanding animal responses to poisonous plants, and preparing animals for new foraging environments. Our research objectives were to determine how variation in 1) flavor concentration (oregano) and 2) dose of gastrointestinal toxin (lithium chloride; LiCl) affected the establishment of CFA in lambs. In a series of experiments feeding lambs ground grains mixed with oregano, we examined how the formation of a CFA to one ground grain, with or without oregano, influenced the consumption of another oregano-flavored grain. We determined that 1) the higher the toxin dose, the stronger the CFA, the greater the generalization of the CFA to a similar feed, and the greater the avoidance of a novel feed; 2) lambs generalized aversions from familiar to novel feeds when both feeds had a flavor in common; and 3) the concentration or intensity of feed flavor apparently did not affect the acquisition or generalization of a CFA, but it did influence the acceptance of a novel feed. Our findings suggest that flavor aversions may be important in the acceptance of harvested or processed feeds.

The concentration of toxins in plants can vary with plant phenology and environmental factors. Changes in toxicity may or may not be associated with changes in plant flavor. These factors may make it difficult for herbivores to avoid toxic plants. We conducted three experiments concerning animals' ability to detect toxins. In experiment 1, we examined whether animals familiar with a tastable toxin on food would adjust intake when toxin concentrations changed. Lambs were initially offered oats with either a low, medium, or high concentration of lithium chloride (LiCl; a salty-tasting toxin). Later, when all lambs were offered oats with the medium concentration of LiCl, lambs that were conditioned with the low LiCl concentration decreased intake while lambs conditioned with the high LiCl concentration increased intake. In experiment 2, we examined flavor aversions formed when animals sampled food with two flavor levels prior to illness. Lambs initially ate the same amount of barley with a low and with a high concentration of either sodium saccharin (a sweet flavor) or aluminum sulfate (a bitter flavor). Lambs then received a mild dose of LiCl. Later, when lambs were offered a choice between barley with either a low or high flavor concentration, lambs preferred the barley with the low flavor concentration. Experiment 3 examined how variable toxicity affected intake when the flavor remained constant. Lambs were initially offered oregano-flavored barley and then received a dose of LiCl. Lambs received either a medium dose, a high dose, or a dose that varied randomly among low, medium, or high amounts of LiCl. Lambs in the medium dose group ate more oregano-flavored barley at the end of the trial than did lambs in either of the other groups.

Generalization occurs when an animal responds to novel stimuli in much the same way it has responded to similar, familiar stimuli. Our objective was to determine whether goats generalized foraging skills from older growth blackbrush (Coleogyne ramosissima) to shrub live oak (Quercus turbinella). Blackbrush (OGBB) has small leaves that are difficult to ingest without the twigs, whereas shrub live oak (SLO) has larger leaves that can be ingested easily without twigs. In Yr 1, 6-mo-old goats and in Yr 2 18-mo-old goats were given experience browsing with OGBB for 0, 10, 20, or 30 d. They were next tested for 5 min/d for 2 d on OGBB, and then tested in the same way on SLO. The number of bites and bite rate increased significantly as goats of both ages gained experience browsing OGBB. Likewise, goats with more experience browsing OGBB took more bites during a 5-min period when foraging on SLO. Thus, goats generalized foraging skills from OGBB to SLO. Plant life form also affected the response of goats. Total intake, intake rate, and bite size were significantly higher for SLO than for OGBB. There was an interaction between level of experience and plant life form. The interaction occurred because goats with no experience browsing OGBB took more bites per 5 min of SLO than of OGBB, whereas goats with increasing levels of experience browsing OGBB generally took more bites of OGBB than of SLO. Controls took more bites per 5 min of SLO than of OGBB, because SLO leaves could be ingested (harvested and chewed) more easily than OGBB twigs.

A lamb's mother and postingestive feedback both influence learning about foods, but their relative importance is unknown. We conducted a study to compare the ingestion of elm (1) by lambs whose mothers avoided elm because ingestion of elm by mother was previously paired with the toxin lithium chloride (LiCl) (M), (2) by lambs who received a mild dose of LiCl after they ingested elm, and whose mothers also avoided elm (M + L), and (3) by lambs who received LiCl after they ingested elm, but whose mothers ate elm avidly (M vs. L); in treatment (4) neither mothers nor lambs were given LiCl (C). In all four treatments, mothers and lambs ate poplar. Each lamb was exposed with its mother for 5 min/day to poplar on days 1,2, 5, 7, 9, and 11 and to elm on days 3, 4, 6, 8, 10, and 12. Following exposures to elm on days 8, 10, and 12, lambs in treatments M + L and M vs. L were given a low dose of LiCl (100 mg/kg body weight orally in a capsule) when they ingested elm to produce an aversion to elm. During the first test after weaning, lambs could choose between elm and poplar. Lambs in C took more bites of elm than did lambs in M, M + L, and M vs. L (13, 2, 2, < 1;P >F=0.003), and they also took a higher percentage of bites from elm (42, 11, 6, 1;P >F < 0.001). The number of bites of poplar did not differ among treatments (M=31, M + L=26, M vs. L=42, C=27;P >F =0.458). During the second test, when lambs were offered only elm, lambs in C took significantly more bites of elm than those in the other treatments, and lambs in M took more bites of elm than lambs in M + L, but not M vs. L (C=38, M=16, M + L=3, M vs. L=6;P >F < 0.001). During the third test, when lambs again could choose between elm and poplar, lambs in C again took more bites of elm (C=14, M=3, M + L=< 1, M vs. L=< 1;P >F=0.034), and they also took a higher percentage of bites from elm (C=26, M=5, M + L=2, M vs. L=2;P >F < 0.001), than did lambs in the other three treatments. The number of bites of poplar did not differ among treatments (M=47, M + L=43, M vs. L=62, C =41;P >F=0.223). We conclude that the response by lambs to the toxin LiCl was more important than was the mother as a social model because lambs that received LiCl avoided elm whether or not the mother ate it.

We investigated whether the odor of garlic can cross the placental barrier from the mother to the fetal environment. Samples of amniotic fluid, allantoic fluid, fetal blood, and maternal blood were collected 0, 50, 100, and 150 min after a pregnant ewe (approximately day 110 of gestation) was gavaged with 6 ml of Egyptian garlic oil. A panel of judges detected (p < 0.05) garlic odor in samples of allantoic fluid, fetal blood, and maternal blood collected 50, 100, and 150 min after the ewe was given garlic and in samples of amniotic fluid collected 100 min after treatment.

We studied lambs' preferences for nonnutritive flavors that were paired with a glucose solution. On the 1st d of the experiment, lambs were offered a saccharin solution flavored with either orange or grape. The following day lambs received a glucose solution containing either orange or grape flavor. Lambs that had received grape and saccharin on d 1 received orange and glucose on d 2, and the reverse was true for the remaining lambs. Conditioning lasted 10 d; odd days were like the first and even days like the second. After conditioning, when lambs were offered a choice between orange- or grape-flavored water without sweeteners, lambs chose the flavor that had been paired with glucose. We also tested the initial hedonic response of naive lambs to the flavor of glucose and saccharin solutions. Lambs exhibited no initial preference. Our results indicate that lambs preferred the flavor that was paired with glucose (calories). Furthermore, results of this study suggest that ruminants may not innately recognize specific chemical constituents in foods or select diets based on initial hedonic value. Rather, learning plays a key role in the formation of dietary preference.

Low nutritional quality and high levels of condensed tannins adversely affect voluntary intake of blackbrush (Coleogyne ramosissima Torr.) by goats. We studied: (1) how consumption of blackbrush or alfalfa pellets by young goats affected their consumption of those foods later in life, and (2) whether previous ingestion of blackbrush or alfalfa pellets affected the excretion of condensed tannins and total phenols from blackbrush in urine and feces, production of proline-rich proteins in saliva, excretion of nitrogen in feces, and mass of the liver, kidneys, parotid glands, and reticuloramen in goats. From 6 to 26 weeks of age, experienced goats were exposed to blackbrush, while inexperienced goats ate alfalfa pellets. Following exposure, both groups were offered older-growth blackbrush twigs (OG) or a choice between OG and current season's blackbrush twigs (CSG). A similar feeding trial was repeated nine months after exposure, and, in addition, both groups were offered a choice between OG ad libitum and alfalfa pellets at six levels of availability. Immediately after exposure, experienced goats ingested 95% more (P < 0.01) OG per unit of body weight than inexperienced goats, but both groups rejected CSG. Nine months after exposure, experienced goats ingested 27% more (P < 0.01) OG than inexperienced goats. Experienced goats ingested 30% more OG than inexperienced goats at every level of alfalfa pellet availability. The fate of condensed tannins and total phenols was similar for both groups, but experienced goats excreted 63% more (P < 0.05) uronic acids per unit of body weight. Neither experienced nor inexperienced goats produced proline-rich proteins in saliva. Experienced goats excreted 32% more (P < 0.01) nitrogen in feces per unit of nitrogen ingested than did inexperienced goats. The mass of the reticulorumen was 30% greater (P < 0.05) for experienced than for inexperienced goats one month after exposure, but did not differ 10 months after exposure. The mass of the liver, kidneys, and parotid glands did not differ between treatments. The results show that experience early in life can have profound and persistent effects on consumption of diets high in chemical defenses and low nutrional quality. The results also suggest that several physiological and morphological factors are involved.

Condensed tannins were isolated from bitterbnish (Purshia tridentata) and blackbrush (Coleogyne ramosissima). Structural analyses showed that both tannins were procyanidins of similar polymer length. The overall stereochemistries at C-3 and C-4, however, differed between the two tannins. These changes in stereochemistry resulted in blackbrush tannins being less preferred than bitterbrush tannins when offered to snowshoe hares (Lepus americanus). It is unlikely that differences in protein-precipitating abilities are the cause for the preference of the bitterbrush over the blackbrush tannins. Instead, we hypothesize that condensed tannins may be depolymerized and absorbed following ingestion. Differences in tannin structure can lead to different depolymerized products and rates of depolymerization, both of which may affect herbivore preferences.

Tall larkspur (Delphinium spp.) is a palatable but toxic poisonous plant in the western United States. The toxins in tall larkspur are diterpenoid alkaloids. We examined the influences of food flavor and postingestive consequences on consumption of a 33% larkspur pellet during 30-min feeding periods for five days using esophageally fistulated cattle that were sham-fed larkspur pellets. Consumption by the sham-fed group was compared to a control group fed alfalfa pellets, and a larkspur group fed only larkspur pellets. Sham-fed cattle did not decrease (P > 0.1) feed consumption compared to controls, indicating no significant difference in food flavor. The larkspur group decreased (P < 0.05) feed consumption by 41% relative to controls and by 31% relative to sham-fed animals (P = 0.08). This reduction in feed consumption indicates the adverse postingestive consequences of tall larkspur ingestion, as the larkspur group apparently developed a conditioned taste aversion to the larkspur pellet. Even though these animals were averted to the pellets, they showed none of the classical signs of intoxication from ingestion of tall larkspur.

We studied the persistence of conditioned taste aversions in sheep and whether or not sheep averted to one shrub species would avoid another shrub species. Three-month-old lambs were averted to the shrub Cercocarpus montanus by pairing its ingestion with lithium chloride (LiCl), a nonlethal gastrointestinal poison. When lambs were yearlings, they were offered C. montanus plants growing in 8-liter pots. During the persistence trial, averted sheep took fewer (P less than .05) bites of C. montanus than controls did (19 vs 64 bites/sheep, respectively). Following the persistence trial, sheep that had been previously averted to C. montanus and consumed C. montanus received LiCl. Complete aversion to C. montanus was re-established in previously averted sheep with a single dose of LiCl. When sheep previously averted to C. montanus were offered potted Amelanchier alnifolia shrubs, averted sheep consumed 21 bites of A. alnifolia but controls consumed 58 bites (P less than .05). This result suggests that previously averted sheep were either more food neophobic than controls or generalized their aversion from C. montanus to A. alnifolia. As with C. montanus, a single dose of LiCl completely averted sheep to A. alnifolia. When sheep grazed a pasture containing C. montanus and A. alnifolia averted sheep took fewer (P less than .05) bites of the two shrubs than controls did (.2% vs 18%). Sheep showed no signs of extinguishing the aversion to either shrub during the grazing season (May-Sept.). Aversive conditioning may be useful to manipulate diet selection of free-ranging livestock.

We investigated the influence of age of exposure to a specific forage on its consumption by young sheep both during and after that exposure. Prior exposure to the shrub Cercocarpus montanus enhanced subsequent consumption (P less than or equal to .05) of that shrub by lambs and yearlings during two experiments in consecutive years. Effects of prior exposure persisted as long as 9 mo without intervening exposure. Age of exposure influenced consumption in lambs both during the period of exposure (P less than or equal to .0001) and during feeding trials following exposure (P less than or equal to .07) in Exp. 1. Lambs exposed at 1 to 5 wk of age, when they are nonruminant and normally are dependent on the dam, browsed less of the time while at the shrub (P less than or equal to .02) and consumed less of the shrub (P less than .02) at a slower rate (P less than or equal to .03) than did lambs exposed at older ages. Lambs exposed at 4 to 8 wk of age, a period corresponding to the transition from nonruminant to ruminant digestion, consumed more shrub following exposure than did lambs exposed at younger or older ages (P less than equal to .05, unprotected least significant differences test). In Exp. 2, which compared naive yearlings with the experienced yearlings exposed as lambs in Exp. 1 the previous year, age of exposure did not influence consumption by yearling sheep following exposure; however, the number of months that had elapsed from the period of exposure varied from 0 to 13 among experimental groups and may have confounded results. Learning of forages appeared poor among nonruminant lambs; data were inconclusive concerning the efficiency of learning of lambs in the transition to ruminant digestion.

Our study determined whether lambs' food preferences developed as a result of exposure to specific foods early in life and whether the presence of the mother affected the development of dietary preferences. In the first experiment, lambs were fed wheat (W) and mountain mahogany (MM) (Cercocarpus montanus) from 50 to 110 d of age. After exposure to W and MM, lambs preferred (P less than .05) W when offered with barley (B) but did not prefer (P greater than .05) MM when offered with serviceberry (SB) (Amelanchier alnifolia), probably because lambs were reluctant to eat mountain mahogany during exposure. Persistence was determined by dividing lambs exposed to W or MM into four groups: 1) continued exposure to W (MM); 2) exposure to B (SB); 3) no further exposure; and 4) exposure to W + B (MM +SB). Preferences for wheat persisted (P less than .05) regardless of the second exposure. In the second experiment, lambs were exposed to W (MM) either alone or with their mothers. Persistence was determined after 21 d of exposure to B (SB). Lambs preferred (P less than .05) W and MM whether they were exposed alone or with their mothers. However, preferences persisted longer in lambs exposed with their mothers. Finally, lambs exposed to either SB or MM with their mothers were later allowed to choose between SB or MM. Lambs strongly preferred (P less than .05) the shrub to which they were exposed early in life.

We studied the ability of lambs to select safe foods in the presence of harmful foods. In a series of feeding experiments, 3- to 6-mo-old lambs were offered a choice between 1) a familiar, safe and a novel, harmful food, 2) a novel, safe and a novel, harmful food and 3) a familiar, harmful and a novel, safe food. All harmful foods were palatable feeds that had been treated with lithium chloride (LiCl), a non-lethal gastrointestinal poison. When lambs ingested a meal composed of a novel food containing LiCl and a familiar, safe food, lambs subsequently avoided the novel food. Lambs avoided the novel food even when the familiar food contained LiCl. The response of lambs varied when lambs were given a choice between a novel, safe food and a novel food containing LiCl. Lambs either avoided the novel food containing LiCl and ingested the safe food or they limited their intake of both novel foods. Their response was dependent on the novelty of the food containing LiCl. When a novel, palatable food contained 2% LiCl, lambs always ate some of the food, even after experiencing illness from ingesting it. Lambs experienced with foods containing LiCl displayed greater food neophobia than lambs naive to LiCl-treated foods. Thus, novelty was the major criterion that lambs used to associate foods with gastrointestinal illness.