HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Villalba, J. J. Articles by Peterson, C. Search for Related Content PubMed PubMed Citation Articles by Villalba, J. J. Articles by Peterson, C.

Phosphorus appetite in sheep: Dissociating taste from postingestive effects^1,2

J. J. Villalba^*,3, F. D. Provenza^*, J. O. Hall and C. Peterson^*

^* Department of Forest, Range and Wildlife Sciences, and Department of Animal, Dairy and Veterinary Sciences, Utah State University, Logan 84322-5230

	Abstract

Top Abstract INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION IMPLICATIONS LITERATURE CITED

 
We hypothesized that lambs discriminate the postingestive effects of P and associate those effects with feed flavor to modify feed choices. Three predictions were tested based on this hypothesis: 1) lambs will modify preference for arbitrary flavors eaten during intraruminal infusions of NaH₂PO₄, 2) changes in preferences are more specific than changes in osmotic load induced by salts of Na; and 3) preference for P is inversely related to the concentration of inorganic P in blood. Thirty lambs were depleted of P by the offer of a P-deficient diet, allocated to 3 groups (10 lambs/group), and conditioned during 3 periods as follows: During conditioning period 1, lambs in each of 3 groups ate a poorly nutritious feed (grape pomace), flavored differently for each group, while water was infused into the rumen. During conditioning periods 2 and 3, lambs again ate grape pomace, with 2 new flavors now paired with infusion of an aqueous solution (126 mmol) of NaCl (conditioning period 2) or NaH₂PO₄ (conditioning period 3), rather than with water. After conditioning, all lambs were offered a choice of the 3 flavors during preference tests immediately after conditioning (period 1) and every 3 wk thereafter (periods 2, 3, and 4). During period 1, when serum inorganic P levels were greatest, lambs preferred flavors paired with water > NaCl > NaH₂PO₄ (P < 0.05). During periods 2 and 3, as inorganic P concentrations decreased in serum, lambs preferred flavors paired with NaH₂PO₄ > NaCl (period 2, P = 0.10; period 3, P = 0.05). Lambs preferred flavors paired with water > NaH₂PO₄ in period 2 (P < 0.001), but those differences disappeared in periods 3 and 4 (P > 0.05). During period 4, lambs preferred flavors paired with NaCl > NaH₂PO₄ (P < 0.10). The estimate of the slope for the linear relationship between intake of flavors paired with NaH₂PO₄ and serum inorganic P was negative (P < 0.0001), whereas estimates of the slopes for the relationships between intake of flavors paired with NaCl or water and serum inorganic P were not different from 0. Thus, preference for P was inversely related to the concentration of serum inorganic P. Our results suggest lambs discriminated among the postingestive effects of NaH₂PO₄, NaCl, and water and associated those effects with specific flavors. Lambs avoided flavors paired with NaH₂PO₄ during periods of P replenishment, and they increased preference for those flavors during periods of P need.

Key Words: choice • feed preference • intake • phosphorus • postingestive learning • sheep

	INTRODUCTION

Top Abstract INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION IMPLICATIONS LITERATURE CITED

 
Phosphorus deficiency is a major problem in many areas of the world. After sodium, P is the most common deficiency for grazing ruminants (Kincaid, 1988). Maintenance of P homeostasis is essential for the proper growth and well-being of young and adult animals (Mulroney et al., 2004). Phosphorus is involved in a wide range of cellular metabolic processes that require high-energy phosphates and in structural roles in cell membranes, bones, and teeth (Ternouth, 1991).

Cows depleted of P develop an avid appetite for P-rich materials, such as bones, which is suppressed by restoring plasma inorganic P concentrations to or above normal values (Denton et al., 1986). Despite this finding, a selective state-dependent appetite for P has been questioned because cattle and sheep have failed to select inorganic P salts as a function of their requirements (Coppock, 1970). Even if an innate P appetite for bones exists in ruminants, it is not known to what extent animals are able to learn about the specific postingestive effects of P and associate those effects with specific flavor cues to modify their feed choices. Lambs acquire preferences for flavors associated with intraruminal infusions of energy (Villalba and Provenza, 1997a) or protein (Villalba and Provenza, 1997b).

A selective preference for the postingestive effects of P can be demonstrated by pairing arbitrary flavor cues in feed with intraruminal infusions of P, such that the flavor and texture of the feed is separated from its postingestive effects (Sclafani, 1995). We conducted an experiment to determine whether lambs 1) associate the postingestive effects of P with different flavors, 2) discriminate the postingestive effects of salts of P from those supplied by salts of Na of similar osmotic loads or from the vehicle (water), and 3) display state-dependent preferences for P such that preference for P is inversely related to the concentration of inorganic P in blood.

	MATERIALS AND METHODS

Top Abstract INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION IMPLICATIONS LITERATURE CITED

 
This study was conducted according to procedures approved by the Utah State University Institutional Animal Care and Use Committee. The study was conducted at the Green Canyon Ecology Center, located at Utah State University in Logan. During the study, 30 commercial, crossbred lambs (4 mo of age) with an initial BW of 41 ± 0.7 kg were individually penned outdoors, under a protective roof, and had free access to trace mineral salt blocks, which did not supply P (Table 1), and fresh water.

Familiarization with the Test Feeds
Lambs were familiarized with a low-quality feed (grape pomace) with moderate amounts of P and containing different flavors (Table 1) known to be easily discriminated by sheep (Villalba and Provenza, 1997c). Throughout familiarization, lambs were given flavored grape pomace at 0730, refusals were collected and weighed at 1130, and all lambs then received 1.5 kg of alfalfa pellets and 300 g of rolled barley grain. No other feed was offered until the next day. Lambs were exposed to each flavor during 3 consecutive periods of 4 d each (from July 22 to August 2, 2004), and the order of flavor exposure was onion-, coconut-, and maple-flavored grape pomace. The pomace was mixed (40 g of flavor/kg of grape pomace) with ground onion powder (Pacific Foods, Kent, WA), coconut, or maple flavors (Lucta USA, Northbrook, IL). Lambs consumed 136 ± 9 g, 186 ± 9 g, and 235 ± 7 g of onion-, coconut-, and maple-flavored grape pomace, respectively.

Initial Preference Test
After the last day of exposure to flavored grape pomace, lambs received onion-, coconut-, and maple-flavored grape pomace simultaneously for 15 min, and intake and preference for each flavor were determined. Preference tests were performed on 2 consecutive days, and the mean of the 2 d was used to assign lambs to 3 groups (group 1, group 2, and group 3; 10 lambs/group) such that lambs with different original flavor preferences occurred equally in all groups.

Phosphorus Depletion
The day after conducting the initial preference tests, we stopped feeding alfalfa pellets and rolled barley, and all lambs were fed a ground (1- to 2-mm particle size) low-P diet consisting of (as-fed basis) 97.2% beet pulp, 2.0% urea, and 0.8% calcium carbonate (3.3 Mcal of DE/kg; 13% CP, NRC, 1985; Table 1). We increased the concentration of Ca in the diet to further reduce plasma inorganic P; P retention by the skeleton increases in response to Ca supplementation (Underwood and Suttle, 1999).

Lambs were offered 2 kg of the low-P diet from 1030 to 1430. At 1430, refusals were collected and weighed. No other feed was offered until the next day. Feeding of the low-P diet began on August 4, 2004, and ended on December 2, 2004.

To determine the level of P deficiency, jugular blood samples were taken from each animal on August 27; September 7, 20, and 28; October 21; November 11; and December 1. Concentrations of inorganic P in the blood are a good indicator of phosphate deficiency in animals (Denton et al., 1986; Miller et al., 1987; Underwood and Suttle, 1999).

Conditioning
After 36 d of exposure to the low-P diet, when the levels of serum inorganic P had decreased to an average of 4.2 mg/dL (Figure 1), lambs were conditioned with the postingestive effects of water, NaCl, and NaH₂PO₄ during 3 consecutive periods of 6 d each between September 9 and September 27, 2004 (Table 2); the sequence of treatments was: water, NaCl, and NaH₂PO₄.

View larger version (14K): [in this window] [in a new window]   Figure 1. Consumption of a low-P basal diet by lambs and changes in serum concentrations of Ca and inorganic P. Lambs were conditioned by pairing of flavored grape pomace with intraruminal infusions of water (d 37 to 42), NaCl (d 43 to 47 and 49), or NaH2PO4 (d 50 to 55). The arrow indicates the day after intraruminal infusions of NaH2PO4 ended. Days to the left of 0 in the bottom graph correspond to the period before lambs began to receive the low-P basal diet; during that period, lambs had free access to a basal diet of alfalfa pellets plus 300 g of rolled barley grain·animal–1·d–1. Values are means for 30 lambs; SE are represented by vertical bars.

The second conditioning period was as described before, but flavors were switched, such that groups 1, 2, and 3 received coconut-, maple-, and onion-flavored grape pomace, respectively. Immediately after beginning to eat the pomace, lambs in all groups were given 200 mL of a water solution with NaCl (126 mmol; Table 2). During this period, lambs consumed 188 ± 4 g, 176 ± 5 g, and 165 ± 6 g of onion-, coconut-, and maple-flavored pomace, respectively.

During the third conditioning period, flavors were switched again, so groups 1, 2, and 3 received maple-, onion-, and coconut-flavored grape pomace, respectively. Immediately after beginning to eat the pomace, lambs in all groups were given 200 mL of a water solution with NaH₂PO₄ (126 mmol; Table 2). This dose is the recommended daily requirement of P for a 40-kg lamb of moderate growth potential (NRC, 1985). During this period, lambs consumed 173 ± 6 g, 175 ± 7 g, and 175 ± 6 g of onion-, coconut-, and maple-flavored pomace, respectively. Thus, NaCl and NaH₂PO₄ treatments delivered the same amounts of Na (2.9 g) and supplied similar osmotic loads. At the end of conditioning, the lambs had a BW of 45 ± 0.8 kg.

The experimental design controlled for several factors. Offering distinctive flavors paired with water, NaCl, or NaH₂PO₄ in 3 different groups controlled for possible flavor effects. Intake of pomace was also similar during different conditioning periods, which controlled for the amount of exposure to the flavors. We delivered NaH₂PO₄ during the last conditioning period to prevent lambs being replenished of P by NaH₂PO₄ infusions before they experienced the water and NaCl treatments. We preferred this approach over a Latin square design to avoid 3 long periods of P depletion before exposure to a new treatment in the Latin square. Because we worked with growing lambs, P deficiencies at different points in time could have diverged across different groups of lambs.

Preference Tests
Preference tests were conducted during 4 consecutive periods. Period 1 was completed the day after conditioning with NaH₂PO₄ ended. All lambs continued to receive the low-P diet every day as described before (Table 2).

Lambs were offered onion-, coconut-, and maple-flavored grape pomace simultaneously for 15 min at 1000, and intake of each fehed was measured (Table 2). No intraruminal infusions of water, NaCl, or NaH₂PO₄ were given during preference tests. From 1030 to 1430, all lambs were offered 2 kg of the low-P basal diet. Percentage preference for flavors associated with each treatment (water, NaCl, or NaH₂PO₄) was calculated as: (intake of a flavored feed/total flavor intake) x 100.

Preference tests were conducted once every 3 wk (periods 2, 3, and 4) on 2 consecutive days in each period (Table 2). Average intake and preference values for each animal were obtained for each period. During d 1 of each period and 1 h before preference tests, jugular blood was obtained from each lamb to determine serum levels of inorganic P and Ca.

Chemical Analyses
Feed Mineral Determinations.
Representative samples of trace mineral blocks, grape pomace, flavored grape pomace, and the low-P basal diet were analyzed for mineral content. Briefly, 0.5 g of each diet material was digested in 9.5 mL of trace mineral grade nitric acid (Fisher Scientific, Pittsburgh, PA) in 50-mL Oak Ridge, Teflon digestion tubes (Nalge Nunc International, Rochester, NY). The samples were digested for 2 h at 90 C.

After digestion, the samples were brought to a final volume of 10 mL with trace mineral-grade nitric acid. Then, 0.5 mL of the digest was added to 9.5 mL of 18.3 Mohm water in 15-mL polypropylene trace metal-free tubes (Elkay, Mansfield, MA). This provided a 5% nitric acid matrix for the analysis, which was matrix-matched for all standard curve and quality control samples. Mineral content analysis was performed using an ELAN 6000 inductively coupled plasma-mass spectrometer (Perkin Elmer, Shelton, CT). Five-point standard curves from 0.01 to 0.50 mg/L were used to quantify the minerals. Sequential 1:10 dilutions were made, using 5% nitric acid, for minerals exceeding the standard curve. Standard curves and quality control samples were analyzed for every 5 samples. National Institute of Standards and Technology standards were analyzed to verify accuracy of the analytical results.

Blood Determinations.
Blood samples were collected by jugular venipuncture. Blood was collected into 10-mL vacuum tubes (Fisher Scientific). After allowing the blood to clot for 1 h, samples were centrifuged for 20 min at 1,100 x g, and the serum was pipetted into trace mineral-free 15-mL centrifuge tubes. Inorganic P and Ca analyses were performed at the local hospital (Logan Regional Hospital) by dry slide techniques of Vitros Products Chemistry (Ortho-Clinical Diagnostics, Rochester, NY),

Statistical Analyses
The effects of group (1, 2, 3), period (1, 2, 3, 4), and treatment (water, NaCl, NaH₂PO₄) or flavor (onion, coconut, maple) received during conditioning on feed intake and on percentage preference during preference tests were assessed using procedures (SAS Inst. Inc., Cary, NC) for a factorial arrangement in a split-split-plot design. Group was the whole-plot, assigned randomly to lambs. Period was the subplot, assigned to repeated trials on each lamb. Treatment (or flavor) was the sub-subplot, assigned to diet components presented to each lamb in each period. Blood inorganic P, measured at the beginning of each period, was a covariate in the analysis.

Intake of the low-P diet and serum inorganic P and Ca values were analyzed as a split-plot design. Group (1, 2, or 3) was the whole-plot, assigned randomly to lambs. Day (low-P diet) or period (serum determinations) was the subplot.

Analyses were computed using the MIXED procedure of SAS. Means were compared using the LSD test. Intake, preference, and blood inorganic P values were subjected, respectively, to square root [sqrt(intake)], arcsine [asin(%preference/100)] and natural logarithmic [ln (inorganic P)] transformations to better meet assumptions of normality and homogeneity of variance.

	RESULTS

Top Abstract INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION IMPLICATIONS LITERATURE CITED

 
Phosphorus Depletion
Intake of the low-P diet did not differ among groups during the 121 d of P restriction (group effect, P = 0.88; group x day interaction, P = 0.72; Figure 1). Averaged across days, lambs in groups 1, 2, and 3 consumed, respectively, 1,228, 1,242, and 1,205 (± 52) g of low-P diet. Intake fluctuated across days, which caused a day effect (P < 0.0001; Figure 1).

Serum inorganic P declined relative to when the basal diet was alfalfa and barley grain when lambs were fed the low-P diet (P < 0.001; Figure 1). Serum inorganic P values increased, relative to all other sampling periods, after the conditioning period with NaH₂PO₄. In contrast, serum Ca values were the lowest after conditioning with NaH₂PO₄ (P < 0.001; Figure 1). All these changes caused a period effect (P < 0.0001), but neither inorganic P (group effect, P = 0.97; group x period interaction, P = 0.92) nor Ca values (group effect, P = 0.94; group x period interaction, P = 0.83) differed among groups.

During P restriction lambs licked and chewed the dirt, and they dug holes in the ground. This behavior was never observed in studies conducted during the past 20 yr at the Green Canyon Ecology Center, when lambs were not subjected to P depletion.

Preference Tests
During initial preference tests before conditioning, lambs did not differ in preference for the flavors associated with each treatment during conditioning (group effect, P = 0.99; group x treatment interaction, P = 0.13; SEM = 5.1). The same was true for intake: 45, 44, and 42 (± 0.6) g, respectively (group effect, P = 0.94; group x treatment interaction, P = 0.64).

Conditioning influenced preference for flavors paired with water, NaCl, and NaH₂PO₄ (intake: treatment effect P = 0.06; treatment x period interaction P = 0.004; preference: treatment effect P = 0.04; treatment x period interaction P = 0.009; Figure 2). Immediately after conditioning, when serum inorganic P levels were the greatest, lambs preferred flavors paired with water > NaCl > NaH₂PO₄ (P < 0.05; Figure 2). During periods 2 and 3, when inorganic P concentrations decreased in serum, lambs ingested (period 2, P < 0.20; period 3, P < 0.10) and preferred (period 2, P = 0.10; period 3, P = 0.05) flavors paired with NaH₂PO₄ over flavors paired with NaCl (Figure 2). Lambs preferred flavors paired with water > NaH₂PO₄ in period 2 (P < 0.001), but those differences disappeared in period 3 (P > 0.05; Figure 2). No differences among treatments were observed in period 4 (P > 0.05) except that preference for flavors paired with NaCl was greater than preferences for flavors paired with NaH₂PO₄ (P < 0.1: Figure 2).

View larger version (18K): [in this window] [in a new window]   Figure 2. Intake and preference for flavored grape pomace by lambs during 15-min preference tests, and changes in serum concentration of inorganic P during 4 periods. Before preference tests, 3 groups with 10 lambs/group were conditioned by pairing of flavored grape pomace with intraruminal infusions of water, NaCl (126 mmol), or NaH2PO4 (126 mmol). Period 1 occurred the day after conditioning, whereas periods 2, 3, and 4 were separated by blocks of 3 wk. Throughout the study, lambs were offered a low-P basal diet. Values are means for 30 (intake, preference) or 10 (serum inorganic P) lambs; SE are represented by vertical bars.

Not all groups responded with the same strength to treatments across periods (group x treatment x period interaction; P < 0.001). For instance, whereas intake of flavors paired with NaH₂PO₄ increased across periods for groups 1 (maple) and 2 (onion; P < 0.05), intake of the flavor paired with NaH₂PO₄ in group 3 (coconut) increased from period 1 to periods 2 and 3, but only from 37 to 50 and 49 g, respectively (P > 0.10), and then it decreased to 28 g during period 4 (P < 0.05). Likewise, intake of coconut-flavored pomace was lower than intake of onion- and maple-flavored pomace across all 4 periods of the study (40 vs. 97 and 62 g, respectively; P < 0.001). All groups of lambs showed this pattern, except that intake of maple- and coconut-flavored grape pomace did not differ for group 2 (62 vs. 54 g; P > 0.05). Coconut was the least preferred flavor by all lambs in the study, across all testing periods. It is likely that doses of NaH₂PO₄ interacted with flavor to reduce preferences for P in group 3 fed coconut-flavored grape pomace.

Serum Inorganic P and Preference Tests
When serum inorganic P was used as a covariate in the analysis with treatment, group, and period in the model, the covariate P level was 0.11 for intake and 0.96 for preference. No significant effects were observed for any interactions of the covariate with group, period, or treatment (P > 0.05). Nevertheless, serum inorganic P and period were strongly correlated (P < 0.001) by design (Figure 2) such that period essentially replaced the covariate. When the model was analyzed without period, there was an effect of serum inorganic P (intake; P = 0.014), and a serum inorganic P x treatment (intake, P = 0.002; preference P = 0.009) and a serum inorganic P x treatment x group interaction (P < 0.001).

Serum inorganic P and preference for flavors paired with NaH₂PO₄ were inversely related. The estimate of the slope for the linear relationship between preference for flavors paired with NaH₂PO₄ and serum inorganic P was significant (intake; –2.26; P < 0.01; preference; –0.10; P = 0.03; Figure 3), indicating an increase in intake and preference for flavors paired with NaH₂PO₄ with the decrease in concentration of serum inorganic P and vice versa. In contrast, estimates of the slopes for the relationships between intake and preference for flavors paired with NaCl or water and serum inorganic P were not different from 0 [intake: –0.01 (P = 0.99) and 0.03 (P = 0.96); preference: 0.06 (P = 0.14) and 0.04 (P = 0.41), respectively; Figure 3]. All groups showed negative slopes for the relationship between consumption of flavors paired with NaH₂PO₄ and serum inorganic P during preference tests. Estimates of the slopes were significant for groups 1 and 2 (P < 0.05), but not group 3 (P > 0.05).

View larger version (15K): [in this window] [in a new window]   Figure 3. Relationship between intake1/2 or arcsin preference for flavored grape pomace by lambs during 15-min preference tests and natural logarithm (ln) of the serum concentration of inorganic P during those tests. Before preference tests, lambs were conditioned by pairing of flavored grape pomace with intraruminal infusions of water, NaCl, or NaH2PO4. Estimate of the slope for the linear relationship between intake1/2 or arcsin preference of flavors paired with NaH2PO4 and natural logarithm of serum inorganic P was negative and significant (intake: P < 0.001; preference: P = 0.03); estimates of the slope for water or NaCl did not differ from 0. Graphs correspond to preference tests of 30 lambs conducted during 4 periods; SE are represented by vertical bars.

	DISCUSSION

Top Abstract INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION IMPLICATIONS LITERATURE CITED

We developed a protocol involving a 3-choice test to discern whether preference for flavors paired with NaH₂PO₄ was different from preference for flavors paired with an infusion of the same osmotic load and Na concentration, but without the benefits/detriments of P (NaCl), and different from an infusion of the vehicle (water). The water control excluded the possibility of nonspecific anorectic effects, which could be due to excesses of P or Na, by allowing lambs to choose a flavored feed not previously paired with P or NaCl. The design also controlled for flavor effects by having 3 groups of lambs each infused with a different substance (NaH₂PO₄, NaCl, water) paired with a different flavor.

Postingestive Effects of P: Avoidance
Immediately after conditioning with NaH₂PO₄, levels of serum inorganic P were greater than when animals were on an alfalfa:barley diet and even above the range reported as normal values for ruminants (Figure 1; 4.5 to 8.5 mg/dL; Field et al., 1984; Kincaid, 1988). Under these conditions of P excess, lambs selected water > NaCl > NaH₂PO₄. Thus, animals avoided P and NaCl at different intensities, suggesting they discriminated among the postingestive effects of water, NaCl, and NaH₂PO₄ and made multiple flavor-feedback associations. Preferences for water were the least variable across declining concentrations of serum inorganic P (Figure 2), further suggesting discrimination among treatments. Sheep also discriminated among 3 flavors associated with ruminal infusions of starch, casein, and water (Villalba and Provenza, 1999), and among the benefits of ingesting sodium bicarbonate, polyethylene glycol, and sodium bentonite when fed diets high in grain, tannins, and oxalates, respectively (Villalba et al., 2006). These data support the concept that sheep sense multiple internal states and modify their foraging behavior accordingly. Sheep also avoid sulfur when requirements for this mineral are met (Hills et al., 1999).

Postingestive Effects of P: Preference
When levels of inorganic P in serum decreased across periods (Figures 1 and 2), lambs increased their intake and preference for flavors previously associated with NaH₂PO₄ infusions, but this increase was never above intake or preference for flavors paired with water (Figure 2). Thus, animals showed a strong avoidance of flavors paired with NaH₂PO₄ when serum inorganic P values were high, and they showed a modest preference for flavors paired with P as serum inorganic P values decreased across periods. We know of 3 explanations for not observing a stronger preference for P. First, serum inorganic P values likely were not low enough to promote a strong preference for flavors associated with NaH₂PO₄ (Figure 1). Cows only develop an avid appetite for old weathered bones as plasma inorganic P drops to approximately 3.1 mg/dL (Blair-West et al., 1992), and osteophagia is rapidly suppressed by increasing plasma inorganic P concentration to or above 5.6 mg/dL (Denton et al., 1986).

Second, we likely delivered too much P during conditioning, thus causing an aversion to the flavor paired with P. In juvenile rats, dietary P restriction causes significant P reabsorption (Woda et al., 2001; Mulroney et al., 2004). Renal adaptation to P reabsorption remained elevated in P-deficient rats despite the subsequent supply of P (Sweeny et al., 1998). If this was the case with lambs, then an enhanced renal P reabsorption likely exacerbated the satiating effects of the doses of P we supplied. Indeed, lambs had above-normal values of serum inorganic P after intraruminal infusions of NaH₂PO₄ (Figure 1).

Third, any aversion to P was likely stronger in some animals than in others. Whereas a dose of a nutrient may be appropriate for the average animal in a group, it will be excessive for certain individuals—even when they belong to the same group or family—due to physiological variability among individuals (Provenza et al., 2003).

More generally, feed flavors associated with gastric infusions of a substance that supplies a benefit stimulate food ingestion, but as concentration of the infusion increases intake is inhibited, presumably due to the satiating effects of the infusion (Provenza, 1995; Ramirez, 1997). Thus, flavor-postingestive consequence learning incorporates both positive reinforcement, which enhances intake and preference of feeds, and anticipated satiety, where animals anticipate the high concentration of nutrients in a feed and as a consequence they decrease intake and preference (Warwick and Weingarten, 1996). During preference tests, lambs likely anticipated the high concentration of P—conditioned by delivery of a single high dose of P during conditioning—previously associated with flavored pomace. Anticipating satiety, based on their previous conditioning, dominated the positive reinforcement of P at the physiological states and dosages we used in our study.

Postingestive Effects of NaCl
Sodium appetite is innate (Spector et al., 1990; Schulkin, 1991). The evolution of Na hunger may have occurred at least in part to regulate intake of other minerals (Schulkin, 2001). Salt licks typically contain other minerals, and the salty taste may attract animals deficient in other minerals (Schulkin, 2001).

An appetite for Na induced by a Ca or P deficiency makes sense evolutionarily only if Ca and P co-occur with Na in nature. Most salt deposits are formed from evaporated sea water, with a mineral composition approximate to that of the early seas, which likely contained high amounts of Na with Ca, but with little P (Tordoff, 1992). Thus, a P deficiency may not have necessarily evolved to induce a Na appetite. In support of this, preference for NaCl in rats is unaltered by deprivation of P or potassium and it even decreases with deprivation of magnesium (Tordoff, 1992). Likewise, in our study lambs did not increase preference for flavors paired with NaCl when challenged with a P deficiency. Rather, their preference for flavors paired with NaCl was lower than preference for flavors paired with water during period 1, and preference for NaCl declined even further when concentration of serum inorganic P diminished during periods 2 and 3 (Figure 2). Slopes of the linear relationship between intake/preference for NaCl and serum inorganic P were not different from 0 (Figure 3). During period 4, preferences for flavors paired with NaCl were slightly greater than preferences for flavors paired with NaH₂PO₄, although they were not greater than preferences for NaCl in period 1 (Figure 2). Lambs had blocks with NaCl available ad libitum in their pens throughout the study. Thus, it is likely that lack of need for Na promoted low preferences for flavored pomace paired with NaCl. Likewise, when the need for NaCl is met, lambs (Villalba and Provenza, 1996) and rats (Revusky et al., 1971) avoid flavors associated with infusions of NaCl.

Intake of the Low-P Basal Diet
Phosphorus-deficient lambs in our study did not reduce feed intake to the degree observed in other studies (e.g., Ternouth and Sevilla, 1990; Ternouth, 1991). Reductions in feed intake have been attributed in part to an impaired rumen function. Phosphorus is essential for ruminal microbial metabolism (Milton and Ternouth, 1984), and a deficiency can reduce digestibility of DM and OM and decrease rates of formation of microbial protein in the rumen (Field et al., 1975; Petri et al., 1988). Nevertheless, the low-P diet we fed in the current study was of very small particle size (1 to 2 mm), and it contained beet pulp, urea, and CaCO₃. The fiber in beet pulp is highly digestible, with 31% cellulose, 21% hemicellulose, and only 2% lignin (Van Soest, 1994). Thus, it is likely that the effects of low P on particle size reduction and on feed intake were not as marked in our study as when roughages of low P concentrations are fed.

Learned Preference for P
Whereas lambs in our study did not show strong preferences for flavors paired with P, there are at least 3 lines of evidence that suggest lambs modified their flavor preferences as a function of the learned postingestive effects of NaH₂PO₄. First, by showing different flavor preferences, animals discriminated among the postingestive effects of NaH₂PO₄, NaCl, and water (Figure 2); this pattern was absent when animals lacked experience with the postingestive effects of NaH₂PO₄, NaCl, or water before conditioning and during initial preference tests. Second, intake and preference of flavored pomace paired with NaH₂PO₄ was the lowest when serum concentrations of inorganic P were the greatest, and there was a gradual increase in intake and preference as serum concentrations dropped across periods. Finally, the slope for the linear relationship between intake and preference for flavors paired with NaH₂PO₄ and serum inorganic P was inverse and significant, whereas the slopes for flavors paired with either NaCl or water and serum inorganic P did not differ from 0 (Figure 3).

Earlier studies have reported that ruminants are unable to recognize the benefits of P supplements. Lactating dairy cows showed highly variable supplement intakes (Coppock et al., 1976), and some animals consumed large amounts of phosphate salts even when they were not needed (Coppock, 1970; Coppock et al., 1972). Likewise, P-deficient cows did not form preferences for inorganic salts of P (Denton et al., 1986; Blair-West et al., 1992). Earlier studies also show osteophagia and allotriophagia in P-deficient sheep that did not form preferences for dicalcium phosphate supplements (Gordon et al., 1954). Findings such as these have led to the conclusion that intake of Ca-P mineral supplements by cattle is related to taste rather than need in livestock (Coppock, 1970), and that sheep select minerals they like rather than those they require (Pamp et al., 1977). Thus, some people recommend feeding minerals in a blended complete feed to ensure adequate mineral consumption (e.g., Muller et al., 1977).

There has been little appreciation for how animals learn flavor-feedback associations (Provenza and Villalba, 2006). Early studies did not take into account past experiences or the contingencies necessary for animals to learn specific flavor-feedback associations. For instance, sulfur-deficient lambs had to discriminate among 4 anions (CO₃^–, PO₄^–, SO₄^–, Cl^–) that all contained Na (Pamp et al., 1977). Animals generalize among substances that share a common flavor (Na), thus making it difficult to discriminate on the basis of similar flavors and excessive feedback from Na. Appropriate learning is likely if animals are offered substances that differ in both flavor and feedback, and if the deficiency is paired with ingesting the substance that rectifies the deficiency. We supplied lambs with distinctive flavors and paired those flavors with different substances (water, NaCl, NaH₂PO₄) while lambs had low serum levels of inorganic P (Figure 1).

Moreover, past studies in which P salts were fed were based on the erroneous assumption that animals eat to meet tabular mineral requirements (Coppock et al., 1976; Pamp et al., 1977). Animals respond most strongly to excesses, deficits, and imbalances (Provenza and Villalba, 2006). When ruminants react to rectify extreme states they may overshoot and undershoot their mineral requirements. Moreover, daily requirements may be exceeded even with a bite per day of a concentrate salt. The daily P requirement of a 40-kg lamb with moderate growth potential is 3.9 g (NRC, 1985), which amounts to only 15 g of NaH₂PO₄.

Recommendations for P requirements may be set at greater levels than animals need (McDowell et al., 1993), and individuals, even within uniform groups, vary in their nutrient requirements (Provenza et al., 2003). Thus, it is not surprising that some animals consume much less or more than others or than recommended requirements. For any individual, when the need for a nutrient is not high, animals likely respond more to the novelty of the taste than to the postingestive effects of the mineral, as they did in early studies (Coppock et al., 1976; Pamp et al., 1977).

The failure to observe preferences for inorganic salts of P may also depend on the type of salt used in the study. For instance, Ca-deficient animals avoid P (Tordoff, 2001). Thus, calcium phosphate salts are not the best choice when exploring specific appetites for Ca or P. In addition, the degree of the mineral deficiency induced in livestock during most early studies likely was not high enough to induce a preference (Coppock et al., 1976) or deficits were assumed (e.g., pastures growing in P-deficient soils) but not measured (Gordon et al., 1954).

Phosphorus deficiency is a major problem in many areas of the world. The ability of P-deficient livestock to form specific preferences for sources of P, and for inorganic salts in particular, is controversial. Moreover, if an appetite for P exists, it is unknown whether it is innate or learned. Our study suggests lambs learn about the specific postingestive effects of P. Selection of P as a function of need may allow individuals to meet their particular P requirements when offered a choice, without concern for supplying adequate amounts of P with single diets, supplements, or pastures, which may not satisfy or exceed particular requirements of individuals within a group. Thus, animals have evolved learning mechanisms to cope with the frequent changes occurring outside and inside their bodies (Provenza and Villalba, 2005).

	IMPLICATIONS

Top Abstract INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION IMPLICATIONS LITERATURE CITED

 
The ability to form specific preferences/avoidances for sources of phosphorus as a function of need implies that strategic distribution of phosphorus-containing supplements in phosphorus-deficient areas may influence grazing pressure and attract herbivores to under-utilized areas, such as those remote from water, enabling more uniform use of forage.

	Footnotes

 
¹ This research was supported by grants from the Utah Agricultural Experiment Station and the Initiative for the Future of Agriculture and Food Systems, USDA (Agreement No. 2001-52103-11215). This paper is published with the approval of the Director, Utah Agricultural Experiment Station, and Utah State University, as journal paper number 7747.

² We thank S. Durham for statistical advice and data analysis. We also thank L. Lisonbee for technical support.

³ Corresponding author: villalba{at}cc.usu.edu

Received for publication November 2, 2005. Accepted for publication March 7, 2006.

	LITERATURE CITED

Top Abstract INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION IMPLICATIONS LITERATURE CITED

 


Blair-West, J. R., D. A. Denton, M. J. McKinley, B. G. Radden, E. H. Ramshaw, and J. D. Wark. 1992. Behavioral and tissue response to severe phosphorus depletion in cattle. Am. J. Physiol. 263:R656–R663.

Coppock, C. E. 1970. Free choice mineral consumption by dairy cattle. Pages 29–35 in Proc. Cornell Nutr. Conf., Ithaca, NY. Cornell Univ., Ithaca, NY.

Coppock, C. E., R. W. Everett, and R. L. Belyea. 1976. Effect of low calcium or low phosphorus diets on free choice consumption of dicalcium phosphate by lactating dairy cows. J. Dairy Sci. 59:571–580.[Abstract/Free Full Text]

Coppock, C. E., R. W. Everett, and W. G. Merrill. 1972. Effect of ration on free choice consumption of calcium-phosphorus supplements by dairy cattle. J. Dairy Sci. 55:245–256.[Abstract/Free Full Text]

Denton, D. A., J. R. Blair-West, M. J. McKinley, and J. F. Nelson. 1986. Physiological analysis of bone appetite (osteophagia). Bioessays 4:40–42.[CrossRef][Medline]

Field, A. C., N. F. Suttle, and D. I. Nisbet. 1975. Effects of diets low in calcium and phosphorus on the development of growing lambs. J. Agric. Sci. (Camb.) 85:435–442.

Field, A. C., J. A. Woolliams, R. A. Dingwall, and C. S. Munro. 1984. Animal and dietary variation in the absorption and metabolism of phosphorus by sheep. J. Agric. Sci. (Camb.) 103:283–291.

Gordon, J. G., D. E. Tribe, and T. C. Graham. 1954. The feeding behaviour of phosphorus-deficient cattle and sheep. Br. J. Anim. Behav. 2:72–74.[CrossRef]

Hills, J., I. Kyriazakis, J. V. Nolan, G. N. Hinch, and J. J. Lynch. 1999. Conditioned feeding responses in sheep to flavoured foods associated with sulphur doses. Anim. Sci. 69:313–325.

Kincaid, R. 1988. Macro elements for ruminants. Page 331 in The Ruminant Animal. Digestive Physiology and Nutrition. D. C. Church, ed. Waveland Press Inc., Prospect Heights, IL.

McDowell, L. R., J. H. Conrad, and F. G. Hembry. 1993. Minerals for grazing ruminants in tropical regions. 2nd ed. U.S.A.I.D. and CBAG, Gainesville, FL.

Miller, W. J., M. W. Neathery, R. P. Gentry, D. M. Blackmon, C. T. Crowe, G. O. Ware, and A. S. Fielding. 1987. Bioavailability of phosphorus from defluorinated and dicalcium phosphates and phosphorus requirements of calves. J. Dairy Sci. 70:1885–1892.[Abstract/Free Full Text]

Milton, J. T. B., and J. H. Ternouth. 1984. The effects of phosphorus upon in vitro microbial digestion. Proc. Aust. Soc. Anim. Prod. 15:472–475.

Muller, L. D., L. V. Schaffer, L. C. Ham, and M. J. Owens. 1977. Cafeteria style free-choice mineral feeder for lactating dairy cows. J. Dairy Sci. 60:1574–1582.[Abstract/Free Full Text]

Mulroney, S. E., C. B. Woda, N. Halaihel, B. Louie, K. McDonnell, J. Schulkin, A. Haramati, and M. Levi. 2004. Central control of renal sodium-phosphate (NaPi-2) transporters. Am. J. Physiol. 286:F647–F652.

NRC. 1985. Nutrient Requirements of Sheep. 6th ed. Natl. Acad. Press, Washington, DC.

Pamp, D. E., R. D. Goodrich, and J. C. Meiske. 1977. Fee choice minerals for lambs fed calcium- or sulfur-deficient rations. J. Anim. Sci. 45:1458–1466.[Abstract/Free Full Text]

Petri, A., H. Müschen, G. Breves, O. Richter, and E. Pfeffer. 1988. Response of lactating goats to low phosphorus intake 2. Nitrogen transfer from rumen ammonia to rumen microbes and proportion of milk protein derived from microbial amino acids. J. Agric. Sci (Camb.) 111:265–271.

Provenza, F. D. 1995. Post-ingestive feedback as an elementary determinant of food preference and intake in ruminants. J. Range Manage. 48:2–17.

Provenza, F. D., and J. J. Villalba. 2006. Foraging in Domestic Vertebrates: Linking the Internal and External Milieu. (In press) Feeding in Domestic Vertebrates: From Structure to Function. V. L. Bels, ed. CABI Publ., Oxfordshire, UK.

Provenza, F. D., J. J. Villalba, L. E. Dziba, S. B. Atwood, and R. E. Banner. 2003. Linking herbivore experience, varied diets, and plant biochemical diversity. Small Rum. Res. 49:257–274.[CrossRef]

Ramirez, I. 1997. Intragastric carbohydrate exerts both intake-stimulating and intake-suppressing effects. Behav. Neurosci. 111:612–622.[CrossRef][Medline]

Revusky, S. H., M. H. Smith, and D. V. Chalmers. 1971. Flavor preference: Effects of ingestion-contingent intravenous saline or glucose. Physiol. Behav. 6:341–343.[CrossRef][Medline]

Schulkin, J. 1991. Sodium Hunger: The Search for a Salty Taste. Cambridge Univ. Press, New York, NY.

Schulkin, J. 2001. Calcium Hunger. Behavioral and Biological Regulation. Cambridge Univ. Press, New York, NY.

Sclafani, A. 1995. How food preferences are learned: Laboratory animal models. Proc. Nutr. Soc. 54:419–427.[CrossRef][Medline]

Spector, A., G. Schwartz, and H. Grill. 1990. Chemospecific deficits in taste detection after selective gustatory deafferentation in rats. Am. J. Physiol. 258:R820–R826.

Sweeny, J. M., H. E. Seibert, C. Woda, J. Schulkin, A. Haramati, and S. E. Mulroney. 1998. Evidence for induction of a phosphate appetite in juvenile rats. Am. J. Physiol. 275:R1358–R1365.

Ternouth, J. H. 1991. The kinetics and requirements of phosphorus in ruminants. Pages 143–151 in Recent Advances on the Nutrition of Herbivores. Proc. 3rd Int. Symp. Nutr. Herb., Serdang, Malaysia.

Ternouth, J. H., and C. C. Sevilla. 1990. The effects of low levels of dietary phosphorus upon the dry matter intake and metabolism of lambs. Aust. J. Agric. Res. 41:175–184.

Tordoff, M. G. 1992. Salt intake of rats fed diets deficient in calcium, iron, magnesium, phosphorus, potassium, or all minerals. Appetite 18:29–41.[CrossRef][Medline]

Tordoff, M. G. 2001. Calcium: Taste, intake, and appetite. Physiol. Rev. 81:1567–1597.[Abstract/Free Full Text]

Underwood, E. J., and N. F. Suttle. 1999. The Mineral Nutrition of Livestock. 3rd ed. CABI Publishing, New York, NY.

Van Soest, P. J. 1994. Nutritional Ecology of the Ruminant. 2nd ed. Cornell Univ. Press, Ithaca, NY.

Villalba, J. J., and F. D. Provenza. 1996. Preference for flavored wheat straw by lambs conditioned with intra-ruminal administrations of sodium propionate. J. Anim. Sci. 74:2362–2368.[Abstract]

Villalba, J. J., and F. D. Provenza. 1997a. Preference for wheat straw by lambs conditioned with intra-ruminal infusions of starch. Br. J. Nutr. 77:287–297.[CrossRef][Medline]

Villalba, J. J., and F. D. Provenza. 1997b. Preference for flavoured foods by lambs conditioned with intra-ruminal administration of nitrogen. Br. J. Nutr. 78:545–561.[CrossRef][Medline]

Villalba, J. J., and F. D. Provenza. 1997c. Preference for flavored wheat straw by lambs conditioned with intra-ruminal infusions of acetate and propionate. J. Anim. Sci. 75:2905–2914.[Abstract/Free Full Text]

Villalba, J. J., and F. D. Provenza. 1999. Nutrient-specific preferences by lambs conditioned with intra-ruminal infusions of starch, casein, and water. J. Anim. Sci. 77:378–387.[Abstract/Free Full Text]

Villalba, J. J., F. D. Provenza, and R. Shaw. 2006. Sheep self-medicate when challenged with illness-inducing foods. Anim. Behav. 71:1131–1139.[CrossRef]

Warwick, Z. S., and H. P. Weingarten. 1996. Flavor-post-ingestive consequence associations incorporate the behaviorally opposing effects of positive reinforcement and anticipated satiety: Implications for interpreting two-bottle tests. Physiol. Behav. 60:711–715.[Medline]

Woda, C., S. E. Mulroney, N. Halaihel, L. Sun, P. V. Wilson, M. Levi, and A. Haramati. 2001. Renal tubular sites on increased phosphate transport and NaPi-2 expression in the juvenile rat. Am. J. Physiol. 280:R1524–R1533.

This article has been cited by other articles:

				J. J. Villalba, F. D. Provenza, and J. O. Hall Learned appetites for calcium, phosphorus, and sodium in sheep J Anim Sci, March 1, 2008; 86(3): 738 - 747. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Villalba, J. J. Articles by Peterson, C. Search for Related Content PubMed PubMed Citation Articles by Villalba, J. J. Articles by Peterson, C.