D. Solà-Oriol, E. Roura and D. Torrallardona

Feed preference in pigs: Effect of cereal sources at different inclusion rates

published online Oct 24, 2008; J Anim Sci

http://jas.fass.orgthe World Wide Web at:

The online version of this article, along with updated information and services, is located on

www.asas.org

by on May 10, 2011. jas.fass.orgDownloaded from

1

Running head: Cereal preference in pigs 1

2

Feed preference in pigs: Effect of cereal sources at different inclusion rates1 3

4

D. Solà-Oriol,* E. Roura,† and D. Torrallardona*

2 5

6

*Animal Nutrition, Institut de Recerca i Tecnologia Agroalimentàries (IRTA), E-43120 7

Constantí, Tarragona, Spain; and †R & D Feed Additives, Lucta SA, E-08170 Montornés del 8

Vallès, Barcelona, Spain 9

10

1This study was supported by the project “Quantitative evaluation of the palatability of

feed and feed ingredients in swine,” Lucta SA, Montornés del Vallès, Barcelona, Spain. The

authors wish to thank the technical support received from IRTA’s laboratory and farm staff.

2Corresponding author: David.Torrallardona@irta.es

Page 1 of 30 Journal of Animal Science

Published Online First on October 24, 2008 as doi:10.2527/jas.2008-0949 by on May 10, 2011. jas.fass.orgDownloaded from

2

ABSTRACT: The palatability of different cereals was studied in 2 two-way choice (preference) 11

experiments using pigs of 56 d of age and 17 kg BW. In Exp. 1, the effect of 24 cereals vs. a 12

common reference diet containing white rice on feed preference in pigs was studied. Pigs were 13

offered a free choice between the reference diet and a diet with the cereal under study for 4 d. 14

Barley, corn (2 sources), wheat, cassava meal, biscuit meal, rye, sorghum, and one source of oats 15

were tested at inclusion rates of 300 and 600 g·kg-1

. Short grain rice (whole, brown, or extruded 16

white), long-grain white rice (raw and cooked), extruded barley, extruded corn, extruded wheat, 17

oats (2 sources), thick rolled oats, cooked oats, and naked oats (raw, extruded, or micronized) 18

were tested at inclusion rates of 150, 300, and 600 g·kg-1

. Relative preference of cereals (% of 19

total feed intake) was affected by type of cereal and by rate of inclusion. The diets containing 20

extruded rice (150 g·kg-1

), extruded naked oats (150, 300, and 600 g·kg-1

), or naked oats (150 and 21

300 g·kg-1

) were preferred (P < 0.05) by pigs to the reference diet. However, the reference diet 22

was preferred (P < 0.05) by pigs to the diets containing 150, 300, and 600 g·kg-1

of cooked long-23

grain rice, oats, or cooked oats, 300 and 600 g·kg-1

of extruded wheat, wheat, corn, sorghum, or 24

unhulled short-grain rice, and 600 g·kg-1

of thick rolled oats, extruded corn, rye, extruded barley, 25

micronized naked oats, barley, cassava, or biscuit meal. Extrusion improved (P < 0.05) 26

preference values for corn and naked oats by pigs, but had no effect on barley, rice, or wheat. In 27

Exp. 2, the pig's preferences for oats and barley were studied using mash and pelleted diets. Diet 28

form did not affect preference in oats diets. However, for barley, greater preference values were 29

obtained when measured in pelleted form compared with mash form. Additionally, direct 2-way 30

choices were also performed between oats and barley diets and between diets presented in mash 31

and pelleted forms. Pigs preferred barley to oats, and preferred diets presented in pelleted to 32

those presented in mash form. In conclusion, cereal type, inclusion rate, and diet form affected 33

Page 2 of 30Journal of Animal Science

by on May 10, 2011. jas.fass.orgDownloaded from

3

feed preference in pigs. Using cereals with greater preference values may contribute to the 34

formulation of more palatable feeds, which enhance feed intake of piglets at critical stages such 35

as weaning time. 36

37

Key words: cereal, palatability, pigs, preference, two-way choice. 38

39

INTRODUCTION 40

Piglets at weaning have difficulty in initiating dry feed consumption. They are removed 41

from the sow and offered dry feed and plain water, which, along with other stress factors (Lalles 42

et al., 2007), results in what has been referred to as post-weaning anorexia. As a consequence, 43

the integrity of the gastrointestinal tract may be compromised (Pluske et al., 1997). Thus, 44

ensuring a fast initiation of feeding immediately after weaning is essential to maintain gut 45

function and piglet growth. Animal taste and smell have evolved to associate beneficial (or 46

nutritious) compounds and detrimental (or toxic) compounds with pleasant and unpleasant 47

sensations, respectively (Goff and Klee, 2006). This may result in innate preferences or aversions 48

for different feeds regardless of previous experiences or post-ingestive mechanisms. Therefore, 49

the newly-weaned pig mostly depends on its oro-nasal sensing systems to evaluate feed, and the 50

most palatable feed ingredients should be used at this stage (Forbes, 1995). 51

Cereals constitute the main energy source in piglet diets and are chosen on the basis of 52

their nutrient composition, ignoring possible differences in palatability. The quantification of pig 53

preferences for different cereals should enable the formulation of highly palatable diets that 54

improve feed intake at weaning. This study aims to quantify the palatability of different cereals 55

with the use of 2-way choice preference tests as proposed by Solà-Oriol (2008). The impact on 56

Page 3 of 30 Journal of Animal Science

by on May 10, 2011. jas.fass.orgDownloaded from

4

feed preference of other factors that have been shown to affect feed intake, such as cereal 57

extrusion and feed pelleting (Medel et al., 1999), has also been studied. 58

59

MATERIALS AND METHODS 60

Animals and Housing 61

All the experiments were conducted at Animal Nutrition Unit, Institut de Recerca i 62

Tecnologia Agroalimentàries (IRTA). The experimental procedures with animals described in 63

this study were approved by IRTA’s Ethical Committee on Animal Experimentation. 64

A total of 912 pigs (Landrace × Pietrain) from 9 production batches in IRTA’s sow herd 65

were used to perform 9 groups of 2-way choice tests. Pigs (456 boars and 456 gilts) initiated the 66

tests at 56-d of age (4 wk post-weaning) with an average BW of 16.8 ± 4.2 kg. The results of our 67

previous study (Sola-Oriol, 2008) indicated that feed preferences did not differ between piglets at 68

weaning and 4 wk after weaning. Therefore, the results of the present study using 56-d-old pigs 69

should be applicable to newly-weaned pigs. Eight of the pig groups were used in Exp. 1 and the 70

last group was used to conduct Exp. 2. For each group of pigs, there were 36 pens, containing 2 71

to 3 pigs each, in 2 weaning rooms (24 and 12 pens, respectively) with automatic, forced 72

ventilation and completely-slatted floors. Each pen was provided with 2 feeders with 2 feeding 73

spaces placed beside each other and an equidistant independent water supply at the opposite 74

wall. Within a pen, the 2 diets compared were offered in the 2 feeders according to a two-way 75

choice protocol. 76

Experimental Design 77

In Exp. 1, each of the 8 pig groups was used to conduct a total of 12 two-way choice 78

comparisons during 3 consecutive 4-d experimental periods (4 comparisons per period). All the 79

Page 4 of 30Journal of Animal Science

by on May 10, 2011. jas.fass.orgDownloaded from

5

2-way choice comparisons were conducted using the same reference diet containing 600 g·kg-1

of 80

white broken rice as the only source of cereal. In each experimental period, 1 of the 2-way 81

comparisons consisted of a “reference diet vs. reference diet” comparison which was used as the 82

control. The remaining 3 two-way comparisons were used to test 3 different cereals. In the first 83

experimental period, the 3 cereals were tested at 300 g·kg-1

for the first 3 groups of pigs and at 84

150 g·kg-1

for the last 5 groups of pigs. In the second period, the inclusion rate of the cereals was 85

increased to 600 and 300 g·kg-1

for the first 3 and last 5 groups of pigs, respectively. Finally, in 86

the third period, the cereals alone (at 1,000 g·kg-1

) were tested against rice on its own for the first 87

3 groups and at 600 g·kg-1

for the last 5 groups of pigs. The same cereal was evaluated during the 88

3 periods for each of the 36 pens in each group of animals, and only the inclusion rate was 89

increased with time. Therefore in Exp. 1, a total of 24 different cereal ingredients were compared 90

to the reference diet with white broken rice. Nine replications for each comparison were 91

obtained. All the diets were presented in mash form. 92

The position of the feeders was not changed during the experiment, and the reference diet 93

was always kept in the same position during the 3 periods. This prevented an adaptation period 94

that would have been necessary to respond to the changes in feeder position and shortened the 95

duration of the measurements. Although it could be argued that this approach would not correct 96

for possible effects of feeder position on preference, a control “reference vs. reference” two-way 97

comparison was included in every group of animals to detect such possible effects. For all the 98

animals, it was confirmed that the pre-starter and starter diets that they had been offered before 99

the start of the trial did not contain rice (reference cereal) or any of the cereals to be tested in 100

each particular group. 101

Page 5 of 30 Journal of Animal Science

by on May 10, 2011. jas.fass.orgDownloaded from

6

The last group of animals (36 pens) was used to conduct Exp. 2, which was divided into 2 102

parts (Exp. 2A and Exp. 2B). In Exp. 2A, 24 pens were used to test barley and oats vs. the 103

reference diet, essentially as described for Exp. 1. Twelve pens were used to compare diets 104

presented in mash form, while the other 12 were used to compare diets presented in pelleted 105

form. Both cereals were tested at the inclusion rates of 200, 400, and 600 g·kg-1

in 3 consecutive 106

4-d periods. Six replications for each comparison were obtained. Exp. 2B was conducted with 107

the last 12 pens. In the first 4-d period, the control comparison “reference diet vs. reference diet” 108

was made with both, mash and pelleted diets (6 pens each). In the second 4-d period, the animals 109

that had been offered pelleted diets during the first period were offered a direct 2-way choice 110

between pelleted diets containing 600 g·kg-1

barley and 600 g·kg-1

oats. The animals that had 111

been offered mash diets were offered the same 2-way choice between barley and oats diets, but 112

presented as mash. Finally, in the third 4-d period, each group of 6 pens was divided into 2 113

groups of 3 pens. One was offered the choice between mash and pelleted barley diets (at 600 114

g·kg-1

), and the other was offered the choice between mash and pelleted oats diets (at 600 g·kg-1

). 115

Six replications for each comparison were obtained, 3 of which had previous experience with 116

pelleted and 3 with previous experience with mash diets. 117

Diets and Feeding 118

A basal diet (Table 1) was formulated to meet or exceed the nutrient requirements for 119

swine (NRC, 1998). All of the cereal ingredients were ground with the same hammer mill using 120

a 3-mm screen. Different cereals were evaluated in diets prepared by replacing the corresponding 121

amount of broken rice from the reference diet with the cereal of interest. To avoid possible 122

interferences on feed palatability caused by ingredients other than the cereals, energy, protein, or 123

AA composition were not corrected. The addition of fat or free AA themselves may affect 124

Page 6 of 30Journal of Animal Science

by on May 10, 2011. jas.fass.orgDownloaded from

7

palatability (Mori et al., 1991a,b; Tabuchi et al., 1991; Mizushige et al., 2007). It was decided 125

not to correct for nutrient composition because it would not have been possible to assess the 126

palatability of the cereals per se, if fat or AA were added to the feed. 127

A short-grain variety of rice obtained from the sieving of polished white rice was taken as 128

the cereal of reference in all the tests. Short grain rice was tested as un-hulled rice (whole 129

unpolished rice), as brown rice (unpolished rice with only the husk of the grain removed), and as 130

extruded white rice (at 150°C, 70 atmosphere pressure, and screw speed of 100 rpm). A long-131

grain variety of rice was also tested as long-grain white rice (broken and small size grains 132

obtained from the sieving of polished white rice) and as cooked long-grain white rice (by thermal 133

steaming in a hydrothermal reactor). Barley was used in raw form (without processing) and as 134

extruded barley (at 145ºC; 70 atmosphere pressure, and 75 rpm). Raw (2 sources) and extruded 135

(at 130ºC, 51 atmosphere pressure, and 75 rpm) corn were tested. Raw (unprocessed) and 136

extruded (at 135ºC, 71 atmosphere pressure, and 75 rpm) wheat were also used. Cassava meal 137

was obtained by grinding cassava roots pelleted at 70ºC. Biscuit meal was obtained by grinding 138

byproducts from the bakery industry such as bread, biscuits, breakfast cereals, cakes, crisps, and 139

snacks that are no longer suitable for human consumption. Rye and sorghum were also tested as 140

raw whole grains. Oats was tested as the raw grain (3 sources) without removing the hull, thick 141

rolled oats (the grain was de-hulled, steamed at 120ºC for 45 min, and then flattened by steel 142

rollers into flakes), and cooked oats (thermal steaming in a hydrothermal reactor at 100ºC for 45 143

min). Finally naked oats, a variety of oats with no outer husk, was also tested raw, extruded (at 144

150ºC, 39 atmosphere pressure, and 10 rpm), and micronized (grinding of grains to particles that 145

are only a few microns in diameter). 146

Page 7 of 30 Journal of Animal Science

by on May 10, 2011. jas.fass.orgDownloaded from

8

The different cereal ingredients tested were analyzed according to AOAC (2000) for DM 147

(method 925.09), ash (method 923.03), crude fiber (method 962.09), ether extract (method 148

920.39), and CP (method 968.06). The results of the analysis are shown in Table 2. 149

In Exp. 1, raw barley, corn (sources 1 and 2), wheat, cassava, biscuit meal, rye, sorghum, 150

and oats (source 1) were tested at inclusion rates of 300 and 600 g·kg-1

against the reference diet 151

and also as pure ingredients (1,000 g·kg-1

) against pure short-grain white rice. Whole rice, brown 152

rice, long-grain white rice, extruded white rice, cooked white rice, extruded barley, extruded 153

corn, extruded wheat, oats (sources 2 and 3), thick rolled oats, cooked oats, naked oats, extruded 154

naked oats, and micronized naked oats were tested at inclusion rates of 150, 300, and 600 g·kg-1

155

against the reference diet. 156

Measurements and Calculations 157

In each pen, feed disappearance from the 2 feeders was measured between d 0 to 4, 4 to 158

8, and 8 to 12. Preference was measured as intake of the test diet expressed as % of total feed 159

intake. The following mathematical equation was used: 160

( ) ( )100

intakediet Referenceintakediet Test

intakediet TestPreference% ×

+= 161

Statistical Analysis 162

Exp. 1. Preference values were compared (between cereals by inclusion rate or between 163

inclusion rates by cereral) by ANOVA using the GLM procedure (SAS Inst. Inc., Cary NC). The 164

mathematical model used was: Yi = µ + αi + εi, where Yi is the preference value for the 165

observations of cereal ingredient (or inclusion rate) i, µ is the general mean of all observations, 166

αi is the effect of the feed cereal source (or of inclusion rate), and ει is the unexplained random 167

error. Means from the ANOVA models were compared using the Tukey's Studentized Range 168

Page 8 of 30Journal of Animal Science

by on May 10, 2011. jas.fass.orgDownloaded from

9

(HSD) Test. Additionally, the preference for each cereal at each inclusion rate was compared to 169

the neutral value of 50% using a Student’s t test. 170

The preferences for diets containing either 300 or 600 g·kg-1

extruded barley, corn, naked 171

oats, white rice, or wheat was also compared to that of the corresponding diets with raw cereal 172

using a Student’s t test. The alpha level used for the determination of significance for all the 173

analysis was 0.05. 174

Exp. 2. Preference values for barley and oats at each inclusion rate (200, 400, and 600 175

g·kg-1

) were analyzed with ANOVA using the GLM procedure of SAS by considering the main 176

effects of diet form, cereal type and their interaction. The mathematical model used was Yij = µ + 177

αi + βj + (αβ)ij + εij, where Yij is the percentage of preference for the observations of diet form i 178

and cereal j, µ is the general mean of all observations, αi is the effect of the diet form (mash or 179

pelleted), βj is the cereal effect (barley or oats), (αβ)ij is the interaction between diet form and 180

cereal, and εij is the unexplained random error. 181

Additionally, Student t-tests were performed to compare: (1) the preference for each 182

cereal at each inclusion rate and each diet form with the neutral value of 50%, (2) ADFI of 183

barley and oats based diets offered as 2-way choice either in mash or pelleted form, and (3) 184

ADFI between mash and pelleted diets offered as 2-way choice according to cereal composition 185

(barley or oats) and previous diet form experience of the pigs (mash or pelleted). 186

187

RESULTS 188

Exp. 1 189

The preference values obtained in the first 3 groups of animals using pure cereals were 190

erratic (i.e., not always agreeing well with those obtained at lower inclusion rates). Taking this 191

Page 9 of 30 Journal of Animal Science

by on May 10, 2011. jas.fass.orgDownloaded from

10

into account and the fact that cereals are always included in a balanced diet under practical 192

conditions, in groups 4 to 8, cereals were not tested in pure form, and, instead, the inclusion rate 193

of 150 g·kg-1

was evaluated. The percentage preference measured for the double reference 194

control did not differ from 50% (overall mean = 50.4; SE = 1.22) in any of the 8 groups of pigs 195

(P-values ranging from 0.29 to 0.94), indicating that the observed preference values may be 196

attributed to dietary changes in cereal ingredients. 197

Preference values for the different cereal ingredients studied relative to the common 198

reference diet are shown in Table 3. Differences among cereals were observed at each of the 199

inclusion rates tested (P < 0.05). Among the cereals tested at inclusion rates of 150 g·kg-1

, the 200

highest preference values were observed for extruded rice, extruded naked oats, and naked oats, 201

and having a preference value greater than 50% indicated that they were preferred to the 202

reference diet (P < 0.05). However, the preference values observed for extruded barley, brown 203

rice, extruded corn, thick rolled oats, extruded wheat, micronized naked oats, long-grain white 204

rice, and unhulled short-grain rice were not different from the reference diet. Finally, the lowest 205

preference values (lower than 50%) were observed for cooked white long-grain rice, oats 206

(sources 2 and 3), and cooked oats (P < 0.05). 207

When included at 300 g·kg-1

, preference values greater than 50% were only observed for 208

raw and extruded naked oats. However, preference values lower than 50% (indicative of 209

preference for the reference diet) were observed for wheat, extruded wheat, cooked long-grain 210

rice, corn (source 1), sorghum, oats (sources 1 to 3), unhulled short-grain rice, and cooked oats 211

(P < 0.05). 212

At the greatest inclusion rate of 600 g·kg-1

, a greater preference value than the neutral 213

value of 50% was only observed for extruded naked oats (P < 0.05). The preference values for 214

Page 10 of 30Journal of Animal Science

by on May 10, 2011. jas.fass.orgDownloaded from

11

extruded rice, brown rice, long-grain white rice, and naked oats were not different from 50%; 215

whereas, thick rolled oats, extruded corn, rye, cooked long-grain rice, extruded barley, 216

micronized naked oats, extruded wheat, barley, cassava, wheat, biscuit meal, corn (sources 1 and 217

2), sorghum, unhulled rice, oats (sources 1 to 3), and cooked oats had preference values below 218

50% (P < 0.05). When the cereals were tested as pure ingredients (1,000 g·kg-1

) using pure white 219

short-grain rice as a reference, the preference values for cassava and barley were greater than 220

50% (P < 0.05), whereas corn (source 1), oats (source 1), and sorghum were lower than 50% (P 221

< 0.05). 222

The effect of the inclusion rate (P < 0.05) was observed regardless of the cereal 223

ingredients. When pure ingredient results were discarded, cereal preference decreased as the rate 224

of inclusion increased, with the exception of extruded naked oats where a greater preference 225

value was observed at the inclusion rate of 300 g·kg-1

than at 150 g·kg-1

(P < 0.05). As mentioned 226

before, the preference values for the pure cereals did not always agree with those obtained at 227

lower inclusion rates. Although preference generally decreased with increasing inclusion rates, 228

greater preference values were obtained with barley, corn (source 2), cassava, biscuit meal, and 229

oats (source 1) at 1,000 g·kg-1

than those cereals included at 600 g·kg-1

(P < 0.05). However, it 230

must be taken into account that pure cereals were compared to pure rice instead of the reference 231

diet, which may explain these observations. 232

The effect of extrusion on the preference of different cereals is shown in Figure 1. 233

Extrusion improved the preference value for corn (P = 0.06 and P < 0.05 at 300 and 600 g·kg-1

, 234

respectively) and naked oats (P < 0.05), but had no effect on barley, rice, or wheat. 235

Exp. 2 236

Page 11 of 30 Journal of Animal Science

by on May 10, 2011. jas.fass.orgDownloaded from

12

The preferences for the double reference control in mash and pelleted form (Exp. 2B, first 237

period) were 44.62 (SE, 3.93) and 46.65% (SE, 4.58%) respectively, which were not different 238

from 50% (P = 0.293 and 0.645, respectively). Preference values for barley and oats by diet form 239

(mash or pellet; Exp. 2A) are presented in Table 4. Greater preference values were observed for 240

barley than for oats at all inclusion rates (P < 0.01), confirming the findings of Exp. 1. A 241

tendency (P < 0.10) for greater preference values was observed in pelleted diets for the inclusion 242

rates of 200 and 400 g·kg-1

. At these inclusion rates, interactions between diet form and cereal 243

type were observed (P = 0.10 and P < 0.05, respectively), indicating that the preference 244

improved when measured in pelleted form for barley but not for oats. 245

Comparisons between barley and oats (Exp. 2B, second period) and mash and pelleted 246

diets (Exp. 2B, third period) are presented in Figures 2 and 3, respectively. Average daily feed 247

intake was greater for barley than for oat-based diets in either mash or pellet form (P < 0.05). 248

Also, intake was greater for pelleted diets than for diets in mash form in both barley and oats (P 249

< 0.05), regardless of previous experience with either mash or pelleted diets. 250

251

DISCUSSION 252

Piglets at weaning have difficulties in initiating dry feed intake. They spend considerable 253

time engaged in exploratory behavior to gather information on the new environment (Gardner et 254

al., 2001), but also spend a lot of time fighting with the new penmates until the new social 255

hierarchy is established (Bruininx et al., 2001). In addition, piglets also seem to have some 256

difficulty in identifying and accepting the new supplies of feed and water. It has been reported 257

that it takes up to 3 wk for piglets to reestablish pre-weaning levels of energy intake, and, 258

meanwhile, their gut integrity can be compromised (Pluske et al., 1997). If a continuous supply 259

Page 12 of 30Journal of Animal Science

by on May 10, 2011. jas.fass.orgDownloaded from

13

of nutrients is maintained after weaning, the detrimental changes in gut structure and function 260

after weaning can be prevented (Kelly et al., 1991; Pluske et al., 1996a,b). Therefore, ensuring a 261

fast initiation of feeding immediately after weaning is essential to maintain gut function and 262

ensure good piglet growth. 263

Cereals are the main source of energy in piglet diets, and they may play an important role 264

in the palatability of feed and its recognition by the weaned piglet. The measurement of 2-way 265

choice preferences against a common reference diet has been proposed as a way of quantifying 266

feed palatability (Solà-Oriol, 2008). The results of our previous study indicated that feed 267

preferences did not differ between newly weaned piglets and pigs at 4 wk post-weaning, 268

provided that the animals from either age group did not have previous experience with any of the 269

2 ingredients tested. Although we used pigs at 4 wk post-weaning, previous experience with the 270

cereals tested was avoided, and therefore, the preference values observed could be applicable to 271

newly weaned pigs. 272

Our results showed a high palatability for cereals not commonly used in piglet feeding 273

such as white rice and naked oats; whereas, a widely used cereal like corn had relatively poor 274

preference values. This is in agreement with the work of Mateos et al. (2006) who showed 275

greater feed intake and weight gain for rice-based diets than for corn-based diets. 276

We observed that pigs have a particularly low preference for oats. This is probably due to 277

its high fiber content, as husk removal (in thick rolled oats) clearly improved the preference. This 278

may also explain the pig's decreased preference for the un-hulled short-grain rice compared to 279

polished white rice. 280

Cereals with relatively high fiber content may increase the bulkiness of the diet and affect 281

intake as a result of the physical capacity of the gastrointestinal tract (Kyriazakis and Emmans, 282

Page 13 of 30 Journal of Animal Science

by on May 10, 2011. jas.fass.orgDownloaded from

14

1995; Whittemore et al., 2001). Mateos et al. (2006) observed that oat hulls reduced feed intake 283

when added to corn-based diets, but had the opposite effect when added to rice-based diets, 284

indicating that fiber only affects palatability above certain rates of inclusion. 285

However, crude fiber content cannot explain the differences observed between other 286

cereal sources. Many plant volatile compounds are produced in plant tissues at specific 287

developmental stages. Although hundreds of volatiles are synthesized at different stages of cereal 288

grain development, only a small subset helps animals recognize the appropriate feeds and avoid 289

poor or dangerous feeds (Goff and Klee, 2006). Studies of flavor preferences and aversions 290

indicate that flavor perception may be linked to the nutritional value associated with the 291

perceived feeds (Gilbertson et al., 1997; Myers et al., 2005). Some of the preference values 292

observed in our study may be explained by the sensory evaluation of the feed, as each cereal 293

ingredient is characterized by a particular volatile profile that can be distinguished by the 294

olfactory system of mammals (Grosch and Schieberle, 1997; Zhou et al., 1999; Sides et al., 2001; 295

Schuh and Schieberle, 2005). 296

Innate preferences may be due to the evolution of taste sensory receptors to recognize 297

specific nutrients or toxic constituents (Mueller et al., 2005). Nutritive characteristics are usually 298

associated with receptors of pleasant tastes: sweet receptors with carbohydrate energy sources, 299

umami with AA, and salty with mineral sources. In contrast, the stimulation of sour and bitter 300

receptors is unpleasant and warns the animal against deteriorated feed or toxic compounds, 301

respectively. This is a potentially important factor when evaluating feed preferences in naive 302

weanling pigs. 303

Some plant volatiles or their precursors may also have antimicrobial or health-promoting 304

activities (Villalba and Provenza, 2005; Goff and Klee, 2006), and as a result, animals may also 305

Page 14 of 30Journal of Animal Science

by on May 10, 2011. jas.fass.orgDownloaded from

15

have developed a preference for them. For example, it is well accepted that grazing herbivores 306

are able to self-medicate by selecting plants with medicinal properties (Villalba and Provenza, 307

2007). 308

It is possible that the high preference values observed in our work for white rice could be 309

related to its health promoting properties. Pluske et al. (2002, 2007) observed that a diet based on 310

white rice improved the resistance of piglets to an E. coli infection, and proposed the use of this 311

cereal as a means of controlling piglets post-weaning diarrhea. Similarly, Mateos et al. (2001) 312

indicated an improved growth in piglets fed cooked rice, but they attributed their results to a 313

greater digestibility and decreased content of non-starch polysaccharides. 314

Feed intake and preference may be regulated by post-ingestive factors as well as sensorial 315

perception. The nutrient composition of the diet has been reported to affect feed intake 316

(Kyriazakis and Emmans, 1995; Whittemore et al., 2001) probably via feedback signals from the 317

gastrointestinal tract that respond to quality and quantity of ingested nutrients. In our studies, the 318

different cereals replaced the white rice from the reference diet on a weight basis without 319

adjusting the nutritional value. Correcting for energy and AA composition was not made as this 320

may affect feed palatability by itself. Fat (Mizushige et al., 2007) and free AA (Mori et al., 321

1991a,b; Tabuchi et al., 1991) have been shown to have an effect on palatability. In addition, 322

under our experimental conditions (Solà-Oriol, 2008), the preference responses are evident as 323

early as the first day of testing, and do not change for at least 10 to 12 d. This indicates a 324

sensorial rather than a post-ingestive (nutritional) response. Nevertheless, feed preference 325

responses to nutritional composition seem to require a period of adaptation longer than the 4 d 326

used in our experiments (Black et al., 1986). Kyriazakis et al. (1991) also reported that although 327

pigs are able to make appropriate choices between diets differing in protein composition to meet 328

Page 15 of 30 Journal of Animal Science

by on May 10, 2011. jas.fass.orgDownloaded from

16

their nutritional needs, they can only do so if they had previous experience with the feeds being 329

offered and had been able to learn about their nutritional consequences. In the current study, pigs 330

did not have previous experience with any of the diets under evaluation, further refuting a 331

possible nutritionally driven choice response. 332

It could be hypothesized that the improved preference observed with extrusion cooking of 333

corn and naked oats was due to an improvement in the availability of their starch and faster 334

glucose release (Medel et al., 1999; Piao et al., 1999) that resulted in an improved palatability. 335

The improvement of feed intake and performance of piglets due to the extrusion of maize 336

supports this hypothesis (Hongtrakul et al., 1998). However, Medel et al. (1999) observed that 337

extrusion of corn improved weight gain of piglets without affecting feed intake, which indicates 338

an effect on nutrient digestibility rather than palatability. Similarly, in our study, the 339

micronization of naked oats had a rather negative effect on preference, despite an expected 340

improvement in nutrient digestibility (Huang et al., 1998). Therefore, there seems to be no clear 341

relationship between nutrient availability and feed preference. Differences in oro-sensorial 342

perception cannot be discarded as it is well known that heat treatments can cause a dramatic 343

change in the profile of volatiles released. 344

Our study on cereal preference was performed using diets in mash form. Considering the 345

fact that diets are often presented as pellets in practice, we performed a 2-way choice to test 346

whether diet form could have an impact on the preference values in oats and barley based diets. 347

While no effect of diet form on the measurement of preference was observed for oats, the 348

preference values for barley were increased in pelleted diets. It seems that pelleting affected 349

palatability of barley in a different way compared with white rice. Therefore, possible 350

interactions between cereal type and technological processing of feed must be considered. 351

Page 16 of 30Journal of Animal Science

by on May 10, 2011. jas.fass.orgDownloaded from

17

Due to the impossibility of testing all cereals against each other, they were compared to 352

the same reference diet containing rice. It could be assumed for a given pair of cereals that the 353

observed relative preferences (vs. rice in the reference diet) would be indicative of which of the 2 354

cereals would be preferred in a direct comparison against each other. In this way, in Exp. 1, it 355

can be expected that barley should be preferred to oats because the preference value for barley 356

was 23% at 600 g·kg-1

; whereas, the preference value for oats ranged between 3 and 8%. The 357

greater intake observed in Exp. 2 for the barley compared with the oat-based diets when offered 358

in a direct 2-way choice supports this approach. 359

The greater preferences observed for pellet over mash feed form for both barley- and oat-360

based diets are in good agreement with the results of Skoch et al. (1983). They observed that 361

when pigs were offered a free choice between corn-based pelleted and mash diets, pigs preferred 362

pelleted diets (85.5 vs. 14.5%, respectively). 363

The nutritional value of feedstuffs is the main criteria for feed formulation, and some 364

aspects such as palatability are often ignored. Our work showed marked differences in pig 365

preference for different cereals. This information may allow nutritionists to choose cereals with a 366

high preference by the piglets, which may contribute to more efficient feed formulation at critical 367

stages such as immediately after weaning. 368

369

LITERATURE CITED 370

AOAC. 2000. Official Methods of Analysis. Assoc. Off. Anal. Chem., Washington, DC. 371

372

Black, J. L., R. G. Campbell, H. Williams, K. J. James, and G. T. Davies. 1986. Simulation of 373

energy and amino acid utilization in the pig. Res. Dev. Agric. 3:121-146. 374

375

Bruininx, E. M. A. M., C. M. C. van der Peet-Schwering, J. W. Schrama, P. F. G. Vereijken, P. 376

C. Vesseur, H. Everts, L. A. den Hartog, and A. C. Beynen. 2001. Individually measured 377

feed intake characteristics and growth performance of group-housed weanling pigs: 378

Page 17 of 30 Journal of Animal Science

by on May 10, 2011. jas.fass.orgDownloaded from

18

Effects of sex, initial body weight, and body weight distribution within groups. J. Anim. 379

Sci. 79:301-308. 380

381

Forbes, J. M. 1995. Voluntary Food Intake and Diet Selection in Farm Animals. CAB 382

International, Wallingford, UK. 383

384

Gardner, J. M., C. F. M. de Lange, and T. M. Widowski. 2001. Belly-nosing in early-weaned 385

piglets is not influenced by diet quality or the presence of milk in the diet. J. Anim. Sci. 386

79:73-80. 387

388

Gilbertson, T. A., D. T. Fontenot, L. D. Liu, H. Zhang, and W. T. Monroe. 1997. Fatty acid 389

modulation of K+ channels in taste receptor cells: Gustatory cues for dietary fat. Am. J. 390

Physiol. Cell Physiol. 41:C1203-C1210. 391

392

Goff, S. A., and H. J. Klee. 2006. Plant volatile compounds: Sensory cues for health and 393

nutritional value? Science 311:815-819. 394

395

Grosch, W., and P. Schieberle. 1997. Flavour of cereal products - a review. Cereal Chem. 74:91-396

97. 397

398

Hongtrakul, K., R. D. Goodband, K. C. Behnke, J. L. Nelssen, M. D. Tokach, J. R. Bergström, 399

W. B. Jr. Nessmith, and I. H. Kim. 1998. The effects of extrusion processing of 400

carbohydrate sources on weanling pig performance. J. Anim. Sci. 76:3034-3042. 401

402

Huang, S. X., W. C. Sauer, M. Pickard, S. Li, and R. T. Hardin. 1998. Effect of micronization on 403

energy, starch and amino acid digestibility in hulls barley for young pigs. Can. J. Anim. 404

Sci. 78:81-87. 405

406

Kelly, D., J. A. Smyth, and K. J. McCracken. 1991. Digestive development of the early-weaned 407

pig. 1. Effect of continuous nutrient supply on the development of the digestive-tract and 408

on changes in digestive enzyme-activity during the 1st week post-weaning. Br. J. Nutr. 409

65:169-180. 410

411

Kyriazakis, I., G. C. Emmans, and C. T. Whittemore. 1991. The ability of pigs to control their 412

protein intake when fed in three different ways. Physiol. Behav. 50:1197-1203. 413

414

Kyriazakis, I. and G. C. Emmans. 1995. The voluntary feed intake of pigs given feeds based on 415

wheat bran, dried citrus pulp and grass meal, in relation to measurements of feed bulk. 416

Br. J. Nutr. 73:191-207. 417

418

Lalles, J. P., P. Bosi, H. Smidt, and C. R. Stokes. 2007. Nutritional management of gut health in 419

pigs around weaning. Proc. Nutr. Soc. 66:240-268. 420

421

Mateos, G. G. et al. 2001. Influence of type of cereal and level of fiber on performance of early-422

weaned piglets. J. Anim. Sci. 79 (Suppl. 1):106-107. (Abstr.) 423

424

Page 18 of 30Journal of Animal Science

by on May 10, 2011. jas.fass.orgDownloaded from

19

Mateos, G. G., F. Martin, M. A. Latorre, B. Vicente, and R. Lazaro. 2006. Inclusion of oat hulls 425

in diets for young pigs based on cooked maize or cooked rice. Anim. Sci. 82:57-63. 426

427

Medel, P., S. Salado, J. C. de Blas, and G. G. Mateos. 1999. Processed cereals in diets for early-428

weaned piglets. Anim. Feed Sci. Tech. 82:145-156. 429

430

Mizushige, T., K. Inoue, and T. Fushiki. 2007. Why is fat so tasty? Chemical reception of fatty 431

acid on the tongue. J. Nutr. Sci. Vitaminol. 53:1-4. 432

433

Mori, M., T. Kawada, T. Ono, and K. Torii. 1991a. Taste preference and protein nutrition and l-434

amino-acid homeostasis in male sprague-dawley rats. Physiol. Behav. 49:987-995. 435

436

Mori, M., T. Kawada, and K. Torii. 1991b. Appetite and taste preference in growing rats given 437

various levels of protein nutrition. Brain Res. Bull. 27:417-422. 438

439

Mueller, K. L., M. A. Hoon, I. Erlenbach, J. Chandrashekar, C.S. Zuker, and N. J. Ryba. 2005. 440

The receptors and coding logic for bitter taste. Nature 434:225-229. 441

442

Myers, K. P., J. Ferris, and A. Sclafani. 2005. Flavour preferences conditioned by postingestive 443

effects of nutrients in preweanling rats. Physiol. Behav. 84:407-419. 444

445

NRC. 1998. Nutrient Requirements of Swine. 10th rev. ed. Natl. Acad. Press, Washington, DC. 446

447

Piao, X. S., B. J. Chae, J. H. Kim, J. Jin, W. T. Cho, and I. K. Han. 1999. Effects of extrusion 448

condition of barley on the growth and nutrient utilization in growing pigs. Asian-Aust. J. 449

Anim. Sci. 12:783-787. 450

451

Pluske, J. R., D. J. Hampson, and I. H. Williams. 1997. Factors influencing the structure and 452

function of the small intestine in the weaned pig: a review. Livest. Prod. Sci. 51:215-236. 453

454

Pluske, J. R., L. Montagne, F. S. Cavaney, B. P. Mullan, D. W. Pethick, and D. J. Hampson. 455

2007. Feeding different types of cooked white rice to piglets after weaning influences 456

starch digestion, digesta and fermentation characteristics and the faecal shedding of beta-457

haemolytic Escherichia coli. Br. J. Nutr. 97:298-306. 458

459

Pluske, J. R., D. W. Pethick, D. E. Hopwood, and D. J. Hampson. 2002. Nutritional influences 460

on some major enteric bacterial diseases of pigs. Nutr. Res. Rev. 15:333-371. 461

462

Pluske, J. R., I. H. Williams, and F. X. Aherne. 1996a. Maintenance of villous height and crypt 463

depth in piglets by providing continuous nutrition after weaning. Anim. Sci. 62:131-144. 464

465

Pluske, J. R., I. H. Williams, and F. X. Aherne. 1996b. Villous height and crypt depth in piglets 466

in response to increases in the intake of cows' milk after weaning. Anim. Sci. 62:145-158. 467

468

Schuh, C., and P. Schieberle. 2005. Characterization of (E,E,Z)-2,4,6-nonatrienal as a character 469

impact aroma compound of oat flakes. J. Agric. Food Chem. 53:8699-8705. 470

Page 19 of 30 Journal of Animal Science

by on May 10, 2011. jas.fass.orgDownloaded from

20

471

Sides, A., K. Robards, S. Helliwell, and M. An. 2001. Changes in the volatile profile of oats 472

induced by processing. J. Agric. Food Chem. 49:2125-2130. 473

474

Skoch, E. R., S. F. Biner, C. W. Deyoe, G. L. Allee, and K. C. Behnke. 1983. Effects of pelleting 475

conditions on performance of pigs fed a corn-soybean meal diet. J. Anim. Sci. 57:922-476

928. 477

478

Solà-Oriol, D. 2008. Quantitative evaluation of the palatability of feed ingredients in swine. PhD 479

Diss. Universitat Autònoma de Barcelona, Bellaterra, Spain. 480

481

Tabuchi, E., T. Ono, H. Nishijo, and K. Torii. 1991. Amino-acid and NaCl appetite, and Lha 482

neuron responses of lysine-deficient rat. Physiol. Behav. 49:951-964. 483

484

Villalba, J. J., and F. D. Provenza. 2005. Foraging in chemically diverse environments: Energy, 485

protein, and alternative foods influence ingestion of plant secondary metabolites by 486

lambs. J. Chem. Ecol. 31:123-138. 487

488

Villalba, J. J., and F. D. Provenza. 2007. Self-medication and homeostatic behaviour in 489

herbivores: learning about the benefits of nature’s pharmacy. Animal 1:1360-1370. 490

491

Whittemore, E. C., G. C. Emmans, B. J. Tolkamp, and I. Kyriazakis. 2001. Tests of two theories 492

of food intake using growing pigs - 2. The effect of a period of reduced growth rate on 493

the subsequent intake of foods of differing bulk content. Anim. Sci. 72:361-373. 494

495

Zhou, M. X., K. Robards, M. Glennie-Holmes, and S. Helliwell. 1999. Analysis of volatile 496

compounds and their contribution to flavor in cereals. J. Agric. Food Chem. 47:3941-497

3953. 498

Page 20 of 30Journal of Animal Science

by on May 10, 2011. jas.fass.orgDownloaded from

21

Table 1. Composition of the basal reference diet, as-fed basis 499

Item Content

Ingredients, g·kg-1

White broken rice 600

Soybean meal, 56% CP1 200

Wheat bran 130

Sunflower oil 30.0

DL-Methionine 1.7

L-Lysine-HCl 4.2

L-Threonine 1.6

L-Tryptophan 0.01

Calcium carbonate 10.1

Dicalcium phosphate 15.0

Salt 3.4

Vitamin and mineral complex2 4.0

Estimated nutrient composition

CP, g·kg-1

179.0

Crude fiber, g·kg-1

19.9

Fat, g·kg-1

43.4

Ash, g·kg-1

52.8

Energy, MJ ME·kg-1

14.1

Calcium, g·kg-1

8.0

Inorganic P, g·kg-1

3.7

Chloride, g·kg-1

3.1

Sodium, g·kg-1

1.5

Methionine, g·kg-1

4.5

Met + Cys, g·kg-1

7.4

Lysine, g·kg-1

12.5

Tryptophan, g·kg-1

2.3

Threonine, g·kg-1

8.1 1HP-300 (Hamlet Protein, Horsens, Denmark). 500

2Provided the following per kg of diet: vitamin A, 10,000 IU; vitamin D3, 2,000 IU; vitamin E, 15 mg; 501

vitamin B1, 1.3 mg; vitamin B2, 3.5 mg; vitamin B12, 0.025 mg; vitamin B6, 1.5 mg; vitamin K3, 2 mg; D-calcium 502

pantothenate, 10 mg; nicotinic acid, 15 mg; biotin, 0.1 mg; folic acid, 0.6 mg; Fe, 80 mg as iron sulfate; Cu, 6 mg as 503

copper sulfate pentahydrate; Co, 0.75 mg as cobalt sulfate; Zn, 150 mg as zinc oxide; Mn, 60 mg as manganese 504

sulfate; I, 0.75 mg as potassium iodate; Se, 0.10 mg as sodium selenite; and ethoxyquin, 150 mg. 505

506

Page 21 of 30 Journal of Animal Science

by on May 10, 2011. jas.fass.orgDownloaded from

22

Table 2. Analyzed nutrient content of the cereal feed ingredients, g·kg-1

507

Cereal Description DM Ash

Crude

fiber Fat CP

Reference White raw short-grain rice 861 5 3 6 71

Rice (short-grain) Unhulled raw 875 37 82 23 66

Brown raw 867 15 8 26 76

White extruded 889 9 2 3 72

Rice (long-grain) White raw 877 9 4 12 89

White cooked 877 15 6 13 74

Barley Raw 909 26 46 18 119

Extruded 897 21 38 20 100

Corn Raw (source 1) 887 13 13 39 82

Raw (source 2) 875 12 15 41 73

Extruded 890 16 14 28 96

Wheat Raw 898 16 28 15 131

Extruded 885 17 17 15 114

Cassava Pellets 880 69 59 7 24

Biscuit meal Bakery byproduct 905 62 13 86 95

Rye Raw 889 16 19 15 97

Sorghum Raw 867 15 12 29 77

Oats Raw (source 1) 895 20 126 52 102

Raw (source 2) 872 26 85 27 97

Raw (source 3) 900 19 91 38 119

Thick rolled (source 2) 889 18 10 53 125

Cooked (source 3) 900 28 92 59 118

Naked oats Raw 878 16 19 67 137

Extruded 898 17 12 79 112

Micronized 907 18 22 77 133

508

Page 22 of 30Journal of Animal Science

by on May 10, 2011. jas.fass.orgDownloaded from

23

Table 3. Preference (% of total feed intake) of different cereal based diets offered in a two-way 509

choice with a common reference diet (Exp. 1)1 510

Inclusion rate, g·kg-1

Cereal Description 150 300

600

1,0002

Rice (short grain) Unhulled raw 37.2defg

18.0efg

*

10.8fgh

*

ND3

Brown raw 56.6abcd

58.8bc

45.6bcd

ND

White extruded 72.0a*

59.5

abc 52.3

b ND

Rice (long-grain) White raw 41.0defg

42.2bcdef

43.5bcd

ND

White cooked 32.8efgh

*

32.6cdefg

*

27.9bcdefg

*

ND

Barley Raw ND

46.6bcde

22.8defgh

*

65.1ab

*

Extruded 64.8abc

56.9bcd

25.8cdefgh

*

ND

Corn Raw (source 1) ND

29.2cdefg

*

11.3fgh

*

32.3bc

*

Raw (source 2) ND

31.4cdefg

8.4gh

*

65.2ab

Extruded 51.1abcde

47.9bcde

33.6bcdef

*

ND

Wheat Raw ND

39.2bcdefg

*

21.7defgh

*

40.6abc

Extruded 47.5bcde

36.3bcdefg

* 23.5defgh

*

ND

Cassava Pellets ND

30.3cdefg

21.8efgh

*

68.6a*

Biscuit meal Bakery by product ND

30.9cdefg

18.4fghi

*

52.2abc

Rye Raw ND

44.5bcde

29.1bcdefg

*

49.1abc

Sorghum Raw ND

26.6defg

*

11.2fgh

*

23.3c*

Oats Raw (source 1) ND

18.1efg

*

3.0h*

30.9

c*

Raw (source 2) 23.1fgh

*

9.9g*

6.6

gh*

ND

Raw (source 3) 11.4h*

10.3

g*

8.3

gh*

ND

Thick rolled (source 2) 50.5abcde

44.9bcde

34.2bcdef

*

ND

Cooked (source 3) 18.5gh

*

12.1fg

*

6.9gh

*

ND

Naked oats Raw 66.6abc

*

65.8ab

*

42.2bcde

ND

Extruded 69.8ab

*

90.1a*

81.8

a*

ND

Micronized 45.6cdef

38.8bcdefg

25.0defgh

*

ND

Pooled SE 4.92 6.09 4.79 7.46

a-iValues in the same column with different letters are different (P < 0.05); n = 9. 511

*Values with this symbol are different from the neutral value of 50% (P < 0.05). A value 512

of 50% indicates that the diet under study and the reference diet are equally preferred. 513 1Cereal based diets were prepared by replacing variable amounts of short-grain raw white 514

rice from the reference diet. 515 22-way choice of pure cereals was performed using pure short-grain raw white rice as 516

reference. 517 3ND: Not determined. 518

Page 23 of 30 Journal of Animal Science

by on May 10, 2011. jas.fass.orgDownloaded from

24

Table 4. Preference (% of total feed intake) of barley and oats based diets offered in a two-way choice with a common reference diet using diets 519

presented in mash or pelleted form (Exp. 2A)1 520

521

Diet form

Mash

Pellet P-values

Inclusion,

g·kg-1

Barley Oats Barley Oats

Pooled

SD Cereal Form Cereal × form

200 60.2b

33.9c 88.2

a*

36.6

c* 18.32 < 0.001 0.054 0.106

400 43.1b

28.8b* 80.8

a* 23.2

b* 22.24 < 0.001 0.092 0.027

600 34.9ab

15.9b*

53.4

a 9.3

b* 21.46 0.002 0.501 0.167

522 a-c

Values in the same row with different letters are different (P < 0.05). 523

*Values with this symbol are different from 50% (P < 0.05). A value of 50% indicates that the diet under study and the reference diet are 524

equally preferred. 525 1Diets were prepared by replacing the corresponding amount of short-grain raw white rice from the reference diet by barley or oats. 526

Means: n = 6. 527

528

529

Page 24 of 30Journal of Animal Science

by on May 10, 2011.

jas.fass.orgD

ownloaded from

25

Figure 1. Percentage of preference of the raw vs. extruded for different cereals tested 530

included at 300 (A) and 600 g·kg-1

(B; Exp. 1). Differences in preference between diets 531

with raw and extruded cereal: *P < 0.05 and †P < 0.10. Values are means (n = 9) with 532

SE represented by vertical bars. 533

534

Page 25 of 30 Journal of Animal Science

by on May 10, 2011. jas.fass.orgDownloaded from

26

300 g·kg-1

Cereal

Barley Corn Naked oats White rice Wheat

% o

f P

refe

ren

ce

0

20

40

60

80

100

Raw

Extruded

*

†

600 g·kg-1

Cereal

Barley Corn Naked oats White rice Wheat

% o

f P

refe

ren

ce

0

20

40

60

80

100

Raw

Extruded

*

*

A

B

535 Manuscript number E-2008-0949; Figure 1 536

Page 26 of 30Journal of Animal Science

by on May 10, 2011. jas.fass.orgDownloaded from

27

Figure 2. Individual ADFI of barley and oats based diets (600 g·kg-1

) offered in 537

a two-way choice; presented either in mash or pellet form (Exp. 2B, 2nd

period)). 538

Differences between barley and oats diets: *P < 0.05. Values are means (n = 6) with SE 539

represented by vertical bars. 540

Page 27 of 30 Journal of Animal Science

by on May 10, 2011. jas.fass.orgDownloaded from

28

Presentation form

Mash Pellet

AD

FI,

g·d

-1

0

200

400

600

800

1000

1200

Barley

Oats

* *

541

Manuscript number E-2008-0949; Figure 2 542

Page 28 of 30Journal of Animal Science

by on May 10, 2011. jas.fass.orgDownloaded from

29

Figure 3. Individual ADFI of mash and pelleted diets offered in a two-way 543

choice according to previous experience with either mash or pelleted diets. Diets 544

contained 600 g·kg-1

of either barley (A) or oats (B; Exp. 2B, 3rd

period). Differences 545

between mash and pelleted diets: *P < 0.05. Values are means (n = 3) with SE 546

represented by vertical bars. 547

Page 29 of 30 Journal of Animal Science

by on May 10, 2011. jas.fass.orgDownloaded from

30

Barley diets

Previous experience

Mash Pellet

AD

FI,

g·d

-1

0

200

400

600

800

1000

1200

1400

1600

Mash

Pellet

* *

Oats diets

Previous experience

Mash Pellet

AD

FI,

g·d

-1

0

200

400

600

800

1000

1200

1400

Mash

Pellet

* *

A

B

548

Manuscript number E-2008-0949; Figure 3 549

550

Page 30 of 30Journal of Animal Science

by on May 10, 2011. jas.fass.orgDownloaded from

Citations http://jas.fass.org#otherarticles

This article has been cited by 1 HighWire-hosted articles:

by on May 10, 2011. jas.fass.orgDownloaded from