Appetite, 1992, 18, 143-154
Effects of lntraruminal Infusions of Sodium Salts on Selection of Hay and Concentrate Foods by Sheep E. A. ENGKU A2AHAN
and J. M. FORBES
Department of Animal Physiology and Nutrition, University of Leeds
Infusion of salts of short-chain fatty acids into the rumen of sheep depresses food intake but the relative importance of the effect of the acids and of osmolality are not fully understood. Three experiments were carried out in which sodium acetate (NaAc) and sodium chloride (NaCl) were infused for 3 h at 4 ~/min and compared with water controls. The effects on the intakes of a concentrate food and hay were examined to study the effect of osmotic and acid manipulation of the rumen on dietary choice. In the first experiment the foods were continuously available while in the second and third the foods were given for 3min at 1.5-h intervals. Salt infusions depressed concentrate intake without significant effect on the intake of hay. The effect of NaAc tended to be greater than that of NaCl and persisted after the end of infusion. The weight of concentrate eaten during the 3-min tests was si~fi~ntly and negatively related to both acetate and sodium concentrations in rumen fluid. The weight of hay eaten was negatively related to concentrate intake. The results demonstrate that the treatments imposed had a differential effect on intake, only depressing that of the concentrate food. They also suggest that, as NaAc has a greater effect than NaCI, the influence on feeding is due both to the osmotic load and the acid effect.
The selection of foods to match dietary intake with requirements is a well-known phenomenon and represents an important factor in the overall determination of food intake. There is considerable evidence that animals of agricultural importance have the ability to select appropriate ratios of energy:protein when given a choice of foods (growing pigs, Kyriazakis et al., 1990; broiler chickens, Sha~~da~ & Forbes, 19W). It is necessary for animals to learn the nutritional properties of the foods on offer if they are to select a mixture appropriate to their needs (Kyriazakis et al., 1990; Shariatmadari & Forbes, 1990). In the sheep, this selection ability is exemplified by the work of Glimp (1971), who observed that growing lambs switched from little discrimination among diets of different digestible energy concentrations when hrst introduced to them, to a preference for high-energy foods as they learned the consequences of eating each one. However, Gordon & Tribe (1951) have demonstrated a case of a failure to select well, in pregnant ewes, and it is likely that the ewes were not given the opportunity
Address correspondence to: Professor J. M. Forbes, Department of Animal Physiology and Nutrition, University of Leeds, Leeds, LS2 9JT, U.K. 019%6663/92/020143+ 12 $03.00/O
0 1992 Academic press Limited
144
E. A. ENGKU AZAHAN
AND J. M. FORBES
to learn the nutritional properties of the range of foods offered. Cropper et al. (1985, 1986) and Hou et al. (1991a) have shown that sheep select proportions of low- and high-protein feeds to give a protein intake matched to their presumed requirements for growth. The motivation for this is strong as sheep are willing to make at least 30 responses in an operant conditioning situation to obtain a food reinforcement in order to obtain a “balanced” diet (Hou et al., 1991b). Volatile fatty acids, principally acetic, propionic and butyric, are major products of fermentation in the rumen and are thus prime candidates for negative feedback signals in the control of food intake. The introduction of salts of volatile fatty acids (VFA), notably acetate, into the rumen depresses food intake (Baile & Forbes, 1974; Forbes and Barrio, 1992). Intraruminal infusion of sodium chloride (NaCl) has also been shown to depress food intake (Ternouth & Beattie, 1971; Carter & Grovum, 1988) and there is considerable doubt as to the specificity of the acetate effect. Irrespective of the mode of action, there have been no reported studies of the effects of sodium salts on selectivity for type of food. The experiments reported here were designed to determine how selection of hay and concentrate are affected by intraruminal infusions of sodium acetate and sodium chloride which have previously been reported to have a depressive effect on the food consumption of animals offered a single diet. Although it would have been useful to have taken blood samples for Na and osmolality analysis, the disturbance to the animals and the risk of infection mediated against this; the literature suggests that most of the effects of salt infusion into the rumen on feed intake is via rumen receptors (Forbes & Barrio, 1992) and it is assumed that the same is true for diet selection. GENERAL METHODS
Animals Six rumen-fistulated crossbred Suffolk sheep (two castrated males and four females) were used in this study The animals had been prepared under halothane/O, general anaesthesia and were allowed at least 3 weeks to recover before the start of the experiments. The three mature sheep were 3 years of age and weighed 87, 66 (female) and 112 kg (castrated male) while the other three animals were 6 months old and had body weights of 46, 40 (females) and 50 kg (castrated male). For 1 week before and throughout the experimental period they were fed ad libitum on a choice of a complete pelleted concentrate food (Hi-D, Lincolnshire Farm Feeds, Grimsby; 50% dried grass, 50% barley with added minerals and vitamins; protein, 14.0%; fibre, 15.5%; oil, 3.0%) and hay chopped into lengths of approximately 50mm (protein, 10.0%; fibre, 40.0%; oil, 2-O%) given in two identical plastic buckets; the positions of the buckets were periodically interchanged to eliminate bias. Water was available at all times but its intake was not monitored. Fresh foods and water were provided each day. Results were analysed using analysis of variance for repeated measures experiments and regression and correlation analysis (SAS, 1982), accounting for between-animal variaton. Intraruminal Infusions and Sampling of Rumen Fluid Experimental fluids were delivered into the rumen via a piece of semi-rigid PVC tubing (internal diameter, 3aOmm) inserted through the rumen cannula bung. This
SALTS AND FOOD CHOICE IN SHEEP
145
tubing protruded some 10 cm into the rumen. For the sampling of rumen fluid, a stiff piece of natural nylon tubing (internal diameter, 4eOmm) was inserted through the rumen cannula bung adjacent to but not touching the infusion tubing. This stiff tubing dipped some 30cm into the rumen. At the rumen end of the tubing small holes were bored to allow ease of flow of rumen fluid up the tubing during sampling. A piece of nylon cloth was wrapped around the holes to sieve out solid rumen particles. A 20ml syringe was used for sampling rumen fluid. Both the infusion and sampling tubings were closed with metal plugs when not in use. All infusions were given over 3 h at a rate of 4ml solution/min. To avoid carry-over effects, at least 2 days elapsed between successive treatments. Chemical Analysis
Rumen fluids were analysed for short-chain fatty acid content using gas-liquid chromatography (GLC, Model 204, Pye Unicam Ltd., Cambridge), and for sodium concentration using a flame photometer (Corning 400, Corning Ltd., Halstead., Essex); osmolality was determined by an osmometer (Advanced Instruments Inc. Massachusetts, USA, Model 3MO).
EXPERIMENT1: CONTINUOUS ACCESS TO Foot
DURING THE EXPERIMENTALPERIOD
In this experiment the effects of two doses of sodium chloride (NaCl) and two doses of sodium acetate (NaAc) on the intake of concentrate and hay were compared with water infusion. Methods
Five rumen-fistulated sheep (all except one of the mature females) were used in an experiment of Latin-square desii The infusion treatments consisted of 2 and 4mM/min of NaAc, 2 and rlr&nin of NaCl and water (control) starting at 1015 hrs. The rate of infusion was not made proportional to the animals’ body weight as it is likely that the younger and lighter animals, being thinner, had a greater volume of rumen contents per unit of body weight than the older and fatter animals but the extent of any differences was not known. Concentrate and hay were available continuously and intakes were recorded hourly during the infusion and for a further 2 h after the infusion stopped. Results
Each of the four salt infusions depressed concentrate intake compared to the control during the 3-h period of infusion, while the intake of hay remained unaffected so that total food was also depressed. There was, however, no significant difference due to the type or dose of salt [Figure l(a)]. In the 2-h period after infusion, the depressive effect of salt infusions on concentrate persisted only in the group treated with the higher dose of NaAc [Figure l(b); P < 0.021. The intake of hay was again unaffected. There were no differences in the responses to treatment due to sex or age of animal in this experiment or those following.
146
E. A. ENGKU
AZAHAN
AND J. M. FORBES
500
400 z q
-
300.
1
Pellets
HOY
(19)
(39)
s I i
200-
100 -
OC
-
C
CL2
Cl4
Ac2
Ac4
C
Cl2
Pellets 400
-
z p
300.
2 al S 2
200.
100
0
04
A&
Ac4
Ac2
Ac4
HOY (IO)
(19)
*
-
C
c12
Cl4
Ac2
Ac4
C
Cl2
a4
(b) FIGURE1. The effect of intraruminal infusion of sodium acetate and sodium chloride on the intake of concentrate pellets and hay (Experiment 1). (a) Infusion period (g/3 h); (b) Postinfusion period (g/2 h) (C) Control; C12, 2 rnhf/min NaCl; C14,4 mM/min NaCl; Ac2, 2m~/min NaAc; Ac4, 4nw/min NaAc. *, significantly different from control (P-c 0.05). Numbers in parentheses in this and subsequent figures are standard errors of treatment means.
EXPERIMENT 2: THREE-MINUTE CHOICEOF FOODSWITH SAMPLINGOF RUMANFLUD AT INTERVALSDURING THE EXPERIMENTAL PERIOD Experiment 1 showed that NaCl and NaAc both significantly depressed the intake of concentrated food while having no effect on hay intake. As both foods were continuously available there was plenty of opportunity for the animals to eat as much as they wished of each food. In this experiment they were allowed access to the two foods simultaneously for only 3 min at intervals of 1.5 h to see whether this would be a more sensitive method of monitoring food preference. In addition, samples of rumen fluid were taken via the fistula for analysis of osmolality and sodium (Na) and acetate (AC) concentrations.
SALTS AND FOOD CHOICE IN SHEEP
147
Methods All six sheep were used in a 3 x 3 Latin-square design with two replicates. The treatments were water, 4 mM/min of NaAc and 4 mM/min of NaCl. Fresh foods were given at 0915 hrs and removed at 1015 hrs. At 1145 hrs, infusion was started on the first sheep. After 1*5h both foods were made available to the sheep for exactly 3 min. At the end of 3 min both food buckets were removed, and this procedure repeated with each of the other sheep in turn. After the final observation, all food buckets were weighed to determine the amount of hay and concentrate consumed within the 3-min period. The animals were then deprived of food until the next feeding time 1.5 h later. The recordings of this short-term feeding were repeated at the end of infusion and 1.5 h post-infusion, starting at 1445 and 1615 hrs, respectively, for the first sheep. After the last recording of the day both hay and concentrate were provided ad libitum. About lOm1 of rumen fluid was sampled from each sheep immediately before each 3-min test. These rumen samples were analysed for sodium (Na) and acetate (AC) concentration and osmolality. Results Both NaCl and NaAc infusions depressed concentrate and food intakes at the end of infusion (both P < O-05), compared to the control (Figure 2), NaAc having a somewhat greater effect. At the postinfusion test NaAc still had a considerable effect on concentrate intake, although this was not significant.
Middle Pellets
End Hay (2)
Pellets
After Pellets
Hay (3)
(9)
+
1 LliEut
:I AC
Hay
(18)
n
C CI AC
(2i
lull .c
C Cl AC
FIGURE 2. Intakes of pellets and hay during 3-min access (g/3min) by sheep infused intraruminally for 3 h with water (C), NaCl (Cl) or NaAc (AC) at 4m~/min (Experiment 2). Middle, test midway through infusion; End, test at end of infusion; After, 1.5 h after the end of infusion. See legend for Figure 1 for further explanation.
148
E. A. ENGKU AZAHAN AND J. M. FORBES
While sodium and acetate concentrations and osmolality in rumen fluid remained relatively constant throughout with water infusion, the NaCl and NaAc groups showed consistent increases in sodium [Figure 3(a)] and osmolality [Figure 3(b)] as the experiment proceeded, the values declining after the end of infusion. The acetate levels [Figure 3(c)] increased significantly during the NaAc treatment and subsequently declined; in the other two groups changes in AC levels were small and variable. Across all tests there were significant negative correlations between concentrate intake (CI, g/3 min) and osmolality (Osm, mOsm/kg) which was, in turn, closely related to sodium concentration (Na, mmoles/l): (pCO.001, ?=O-80)
CI=276-0.58sm Osm = 224.9 + 23.1 Na
(p
Effects of lntraruminal Infusions of Sodium Salts on Selection of Hay and Concentrate Foods by Sheep E. A. ENGKU A2AHAN
and J. M. FORBES
Department of Animal Physiology and Nutrition, University of Leeds
Infusion of salts of short-chain fatty acids into the rumen of sheep depresses food intake but the relative importance of the effect of the acids and of osmolality are not fully understood. Three experiments were carried out in which sodium acetate (NaAc) and sodium chloride (NaCl) were infused for 3 h at 4 ~/min and compared with water controls. The effects on the intakes of a concentrate food and hay were examined to study the effect of osmotic and acid manipulation of the rumen on dietary choice. In the first experiment the foods were continuously available while in the second and third the foods were given for 3min at 1.5-h intervals. Salt infusions depressed concentrate intake without significant effect on the intake of hay. The effect of NaAc tended to be greater than that of NaCl and persisted after the end of infusion. The weight of concentrate eaten during the 3-min tests was si~fi~ntly and negatively related to both acetate and sodium concentrations in rumen fluid. The weight of hay eaten was negatively related to concentrate intake. The results demonstrate that the treatments imposed had a differential effect on intake, only depressing that of the concentrate food. They also suggest that, as NaAc has a greater effect than NaCI, the influence on feeding is due both to the osmotic load and the acid effect.
The selection of foods to match dietary intake with requirements is a well-known phenomenon and represents an important factor in the overall determination of food intake. There is considerable evidence that animals of agricultural importance have the ability to select appropriate ratios of energy:protein when given a choice of foods (growing pigs, Kyriazakis et al., 1990; broiler chickens, Sha~~da~ & Forbes, 19W). It is necessary for animals to learn the nutritional properties of the foods on offer if they are to select a mixture appropriate to their needs (Kyriazakis et al., 1990; Shariatmadari & Forbes, 1990). In the sheep, this selection ability is exemplified by the work of Glimp (1971), who observed that growing lambs switched from little discrimination among diets of different digestible energy concentrations when hrst introduced to them, to a preference for high-energy foods as they learned the consequences of eating each one. However, Gordon & Tribe (1951) have demonstrated a case of a failure to select well, in pregnant ewes, and it is likely that the ewes were not given the opportunity
Address correspondence to: Professor J. M. Forbes, Department of Animal Physiology and Nutrition, University of Leeds, Leeds, LS2 9JT, U.K. 019%6663/92/020143+ 12 $03.00/O
0 1992 Academic press Limited
144
E. A. ENGKU AZAHAN
AND J. M. FORBES
to learn the nutritional properties of the range of foods offered. Cropper et al. (1985, 1986) and Hou et al. (1991a) have shown that sheep select proportions of low- and high-protein feeds to give a protein intake matched to their presumed requirements for growth. The motivation for this is strong as sheep are willing to make at least 30 responses in an operant conditioning situation to obtain a food reinforcement in order to obtain a “balanced” diet (Hou et al., 1991b). Volatile fatty acids, principally acetic, propionic and butyric, are major products of fermentation in the rumen and are thus prime candidates for negative feedback signals in the control of food intake. The introduction of salts of volatile fatty acids (VFA), notably acetate, into the rumen depresses food intake (Baile & Forbes, 1974; Forbes and Barrio, 1992). Intraruminal infusion of sodium chloride (NaCl) has also been shown to depress food intake (Ternouth & Beattie, 1971; Carter & Grovum, 1988) and there is considerable doubt as to the specificity of the acetate effect. Irrespective of the mode of action, there have been no reported studies of the effects of sodium salts on selectivity for type of food. The experiments reported here were designed to determine how selection of hay and concentrate are affected by intraruminal infusions of sodium acetate and sodium chloride which have previously been reported to have a depressive effect on the food consumption of animals offered a single diet. Although it would have been useful to have taken blood samples for Na and osmolality analysis, the disturbance to the animals and the risk of infection mediated against this; the literature suggests that most of the effects of salt infusion into the rumen on feed intake is via rumen receptors (Forbes & Barrio, 1992) and it is assumed that the same is true for diet selection. GENERAL METHODS
Animals Six rumen-fistulated crossbred Suffolk sheep (two castrated males and four females) were used in this study The animals had been prepared under halothane/O, general anaesthesia and were allowed at least 3 weeks to recover before the start of the experiments. The three mature sheep were 3 years of age and weighed 87, 66 (female) and 112 kg (castrated male) while the other three animals were 6 months old and had body weights of 46, 40 (females) and 50 kg (castrated male). For 1 week before and throughout the experimental period they were fed ad libitum on a choice of a complete pelleted concentrate food (Hi-D, Lincolnshire Farm Feeds, Grimsby; 50% dried grass, 50% barley with added minerals and vitamins; protein, 14.0%; fibre, 15.5%; oil, 3.0%) and hay chopped into lengths of approximately 50mm (protein, 10.0%; fibre, 40.0%; oil, 2-O%) given in two identical plastic buckets; the positions of the buckets were periodically interchanged to eliminate bias. Water was available at all times but its intake was not monitored. Fresh foods and water were provided each day. Results were analysed using analysis of variance for repeated measures experiments and regression and correlation analysis (SAS, 1982), accounting for between-animal variaton. Intraruminal Infusions and Sampling of Rumen Fluid Experimental fluids were delivered into the rumen via a piece of semi-rigid PVC tubing (internal diameter, 3aOmm) inserted through the rumen cannula bung. This
SALTS AND FOOD CHOICE IN SHEEP
145
tubing protruded some 10 cm into the rumen. For the sampling of rumen fluid, a stiff piece of natural nylon tubing (internal diameter, 4eOmm) was inserted through the rumen cannula bung adjacent to but not touching the infusion tubing. This stiff tubing dipped some 30cm into the rumen. At the rumen end of the tubing small holes were bored to allow ease of flow of rumen fluid up the tubing during sampling. A piece of nylon cloth was wrapped around the holes to sieve out solid rumen particles. A 20ml syringe was used for sampling rumen fluid. Both the infusion and sampling tubings were closed with metal plugs when not in use. All infusions were given over 3 h at a rate of 4ml solution/min. To avoid carry-over effects, at least 2 days elapsed between successive treatments. Chemical Analysis
Rumen fluids were analysed for short-chain fatty acid content using gas-liquid chromatography (GLC, Model 204, Pye Unicam Ltd., Cambridge), and for sodium concentration using a flame photometer (Corning 400, Corning Ltd., Halstead., Essex); osmolality was determined by an osmometer (Advanced Instruments Inc. Massachusetts, USA, Model 3MO).
EXPERIMENT1: CONTINUOUS ACCESS TO Foot
DURING THE EXPERIMENTALPERIOD
In this experiment the effects of two doses of sodium chloride (NaCl) and two doses of sodium acetate (NaAc) on the intake of concentrate and hay were compared with water infusion. Methods
Five rumen-fistulated sheep (all except one of the mature females) were used in an experiment of Latin-square desii The infusion treatments consisted of 2 and 4mM/min of NaAc, 2 and rlr&nin of NaCl and water (control) starting at 1015 hrs. The rate of infusion was not made proportional to the animals’ body weight as it is likely that the younger and lighter animals, being thinner, had a greater volume of rumen contents per unit of body weight than the older and fatter animals but the extent of any differences was not known. Concentrate and hay were available continuously and intakes were recorded hourly during the infusion and for a further 2 h after the infusion stopped. Results
Each of the four salt infusions depressed concentrate intake compared to the control during the 3-h period of infusion, while the intake of hay remained unaffected so that total food was also depressed. There was, however, no significant difference due to the type or dose of salt [Figure l(a)]. In the 2-h period after infusion, the depressive effect of salt infusions on concentrate persisted only in the group treated with the higher dose of NaAc [Figure l(b); P < 0.021. The intake of hay was again unaffected. There were no differences in the responses to treatment due to sex or age of animal in this experiment or those following.
146
E. A. ENGKU
AZAHAN
AND J. M. FORBES
500
400 z q
-
300.
1
Pellets
HOY
(19)
(39)
s I i
200-
100 -
OC
-
C
CL2
Cl4
Ac2
Ac4
C
Cl2
Pellets 400
-
z p
300.
2 al S 2
200.
100
0
04
A&
Ac4
Ac2
Ac4
HOY (IO)
(19)
*
-
C
c12
Cl4
Ac2
Ac4
C
Cl2
a4
(b) FIGURE1. The effect of intraruminal infusion of sodium acetate and sodium chloride on the intake of concentrate pellets and hay (Experiment 1). (a) Infusion period (g/3 h); (b) Postinfusion period (g/2 h) (C) Control; C12, 2 rnhf/min NaCl; C14,4 mM/min NaCl; Ac2, 2m~/min NaAc; Ac4, 4nw/min NaAc. *, significantly different from control (P-c 0.05). Numbers in parentheses in this and subsequent figures are standard errors of treatment means.
EXPERIMENT 2: THREE-MINUTE CHOICEOF FOODSWITH SAMPLINGOF RUMANFLUD AT INTERVALSDURING THE EXPERIMENTAL PERIOD Experiment 1 showed that NaCl and NaAc both significantly depressed the intake of concentrated food while having no effect on hay intake. As both foods were continuously available there was plenty of opportunity for the animals to eat as much as they wished of each food. In this experiment they were allowed access to the two foods simultaneously for only 3 min at intervals of 1.5 h to see whether this would be a more sensitive method of monitoring food preference. In addition, samples of rumen fluid were taken via the fistula for analysis of osmolality and sodium (Na) and acetate (AC) concentrations.
SALTS AND FOOD CHOICE IN SHEEP
147
Methods All six sheep were used in a 3 x 3 Latin-square design with two replicates. The treatments were water, 4 mM/min of NaAc and 4 mM/min of NaCl. Fresh foods were given at 0915 hrs and removed at 1015 hrs. At 1145 hrs, infusion was started on the first sheep. After 1*5h both foods were made available to the sheep for exactly 3 min. At the end of 3 min both food buckets were removed, and this procedure repeated with each of the other sheep in turn. After the final observation, all food buckets were weighed to determine the amount of hay and concentrate consumed within the 3-min period. The animals were then deprived of food until the next feeding time 1.5 h later. The recordings of this short-term feeding were repeated at the end of infusion and 1.5 h post-infusion, starting at 1445 and 1615 hrs, respectively, for the first sheep. After the last recording of the day both hay and concentrate were provided ad libitum. About lOm1 of rumen fluid was sampled from each sheep immediately before each 3-min test. These rumen samples were analysed for sodium (Na) and acetate (AC) concentration and osmolality. Results Both NaCl and NaAc infusions depressed concentrate and food intakes at the end of infusion (both P < O-05), compared to the control (Figure 2), NaAc having a somewhat greater effect. At the postinfusion test NaAc still had a considerable effect on concentrate intake, although this was not significant.
Middle Pellets
End Hay (2)
Pellets
After Pellets
Hay (3)
(9)
+
1 LliEut
:I AC
Hay
(18)
n
C CI AC
(2i
lull .c
C Cl AC
FIGURE 2. Intakes of pellets and hay during 3-min access (g/3min) by sheep infused intraruminally for 3 h with water (C), NaCl (Cl) or NaAc (AC) at 4m~/min (Experiment 2). Middle, test midway through infusion; End, test at end of infusion; After, 1.5 h after the end of infusion. See legend for Figure 1 for further explanation.
148
E. A. ENGKU AZAHAN AND J. M. FORBES
While sodium and acetate concentrations and osmolality in rumen fluid remained relatively constant throughout with water infusion, the NaCl and NaAc groups showed consistent increases in sodium [Figure 3(a)] and osmolality [Figure 3(b)] as the experiment proceeded, the values declining after the end of infusion. The acetate levels [Figure 3(c)] increased significantly during the NaAc treatment and subsequently declined; in the other two groups changes in AC levels were small and variable. Across all tests there were significant negative correlations between concentrate intake (CI, g/3 min) and osmolality (Osm, mOsm/kg) which was, in turn, closely related to sodium concentration (Na, mmoles/l): (pCO.001, ?=O-80)
CI=276-0.58sm Osm = 224.9 + 23.1 Na
(p
