Evaluation of oligosaccharide addition to dog diets:

in¯uences on nutrient digestion and microbial

populations

$

J.A. Strickling

a

, D.L. Harmon

a,*

, K.A. Dawson

a

, K.L. Gross

b

a

Department of Animal Sciences, University of Kentucky, Lexington, KY 40546-0215, USA

b

Hill's Pet Nutrition Topeka, Topeka, KS 66601-1658, USA

Received 4 August 1999; received in revised form 6 March 2000; accepted 8 June 2000

Abstract

Seven adult mixed breed female dogs (17:4 2:9 kg) surgically ®tted with ileal T-cannulas were

used in a 4 7 incomplete Latin square design experiment to evaluate oligosaccharide

supplementation on dry matter (DM), nitrogen (N), ammonia, volatile fatty acid (VFA), bacteria,

blood glucose concentrations, ileal pH, and fecal consistency. Fructooligosaccharide (FOS),

mannanoligosaccharide (MOS), and xylooligosaccharide (XOS) were added at 5 g/kg of diet DM.

There were no differences in DM digestibility, diet or fecal N, N digestibility, ileal or fecal

ammonia, fecal consistency, ileal bacteria colony forming units, blood glucose, or ileal pH. Ileal

butyrate proportion tended to be greater (P ˆ 0:07) in the control diet (0.076 of total VFA)

compared with the oligosaccharide supplemented diets and lower (P ˆ 0:07) for the MOS diet

compared with the FOS and XOS diets. Ileal propionate tended to be higher (P ˆ 0:09) in MOS

(0.198 of total VFA) than FOS and XOS. Fecal bi®dobacteria numbers were unaffected by dietary

treatment. Fecal Clostridium perfringens tended to be lower (P ˆ 0:09) in MOS when compared to

FOS and XOS. Oligosaccharides had relatively minor effects on bacterial growth in the large

intestine and VFA proportions in the small intestine of the canine. For oligosaccharide feeding to

cause microbial changes in the canine greater amounts of oligosaccharide may be required, or it

may require application in select dietary situations. # 2000 Elsevier Science B.V. All rights

reserved.

Keywords: Canine; Digestion; Oligosaccharides; Fermentation; Bacteria

Animal Feed Science and Technology

86 (2000) 205±219

$

Approved by the director of the Kentucky Agricultural Experimental Station as publication 99-07-75.

*

Corresponding author. Tel.: ‡1-859-257-7516; fax: ‡1-859-257-3412.

E-mail address: [e-mail] (D.L. Harmon).

0377-8401/00/$ ± see front matter # 2000 Elsevier Science B.V. All rights reserved.

PII: S 0 3 7 7 - 8 4 0 1 ( 0 0 ) 0 0 1 7 5 - 9

1. Introduction

Oligosaccharides are naturally occurring carbohydrates that are commonly found in

plants. Their chemical and physical properties may vary as a function of structure which

can be linear or branched, linkages can be a or b, and the number and type of monomers

can also vary. Most oligosaccharides are very similar to non-starch polysaccharides

except that they are soluble in water and physiological ¯uids. The small intestine does not

contain the digestive enzymes required to break down these structures; therefore, these

carbohydrates will enter the large intestine in an intact form.

The amount and type of fermentable carbohydrate that arrives in the human colon is

one of the primary factors that limits the growth of the resident bacterial population

(Cummings and Macfarlane, 1991). The bacteria that can most rapidly degrade and use

the digesta will proliferate beyond the others (Cummings and Macfarlane, 1991). Bacteria

comprise 0.40±0.50 of the fecal solids in humans on a Western diet (Stephen and

Cummings, 1980).

While the biological effects of fructooligosaccharides have been varied, often they

have been shown to decrease constipation, blood pressure, blood lipids, and cholesterol in

humans (Hidaka et al., 1986). Fructooligosaccharides fed to humans also decrease mean

fasting blood glucose, mean serum cholesterol, and LDL cholesterol in diabetic subjects

(Yamashita et al., 1984). Rats fed a diet containing oligofructose had a signi®cant

reduction in total mass of body fat as well as decreased lipidemia and a decreased

intrahepatic lipid concentration (Delzenne et al., 1993).

Mannanoligosaccharides have been suggested to adsorb high proportions of pathogens

including certain Salmonella and Clostridium species, as well as Escherischia coli K:88

(Newman, 1994). Spring et al. (2000) challenged 3-day-old chicks with S. dublin and

reported lower cecal colonization at 10 days post challenge for chicks fed

mannanoligosaccharides.

Xylooligosaccharides, as with other oligosaccharides, are not hydrolyzed in the small

intestine and reach the colon in an intact form. They like fructooligosaccharides can then

be used as a substrate for bi®dobacteria (Bunce et al., 1995a; Okazaki et al., 1990).

Manipulating intestinal micro¯ora through supplementation of oligosaccharides has

the potential to alter the colonization of enteric bacteria, thereby, affecting the overall

health of the host. The objective of this experiment was to evaluate ileal and total tract

effects on nutrient digestion and microbial populations when oligosaccharides are added

to dog diets.

2. Materials and methods

2.1. Dogs

Seven adult, mixed breed female dogs with body weights of 17:4 2:9 kg were used in

this experiment to evaluate the effect of oligosaccharide addition to diets on nutrient

digestion and microbial populations. All seven dogs had been surgically ®tted with a

polyvinyl chloride T-cannula 6±10 cm from the ileal±cecal junction (Walker et al., 1994).

206

J.A. Strickling et al. / Animal Feed Science and Technology 86 (2000) 205±219

All procedures described herein were approved by the Institutional Animal Care and Use

Committee.

The animals were located in the Division of Laboratory Animal Resources at the

University of Kentucky, Lexington, in an environmentally controlled room at 228C with a

14 h:10 h light:dark schedule. The dogs were exercised daily for a minimum of 25 min

except on blood sampling and ileal collection days. Each dog was housed in an individual

cage. Five dogs were in stainless steel cages (1:2 m 1:8 m) with a raised step

(1:01 m 0:46 m). The elevated ¯oor and the step were diagonally grated, coated, and

removable. The sixth dog was housed in a cage (0:78 m 1:8 m) with chain link sides

and ceiling. The ®berglass ¯oor and raised step (0:78 m 0:51 m) were solid and not

grated with a small circular drain opening in the center. The seventh dog was in a

(1:3 m 0:90 m) crate without a step. The ¯oor was plastic coated and diagonally grated.

The sidewalls were made of stainless steel.

2.2. Feeding and treatments

All diets (Table 1) were prepared by the Hil's Science and Technology Center (Topeka,

KS). Diets were formulated using current guidelines for dogs (American Association of

Table 1

Composition of the experimental diets

Ingredients

Concentration (g/kg)

Maize

400

Poultry meal

200

Soybean meal

150

Maize starch

91.05

Choice white grease

73

Corn gluten

50

Soy oil

10

Flavor

a

10

Salt

6

Chromic oxide

2

Vitamins, minerals

a

2.75

Ethoxyquin

0.2

Cornstarch or oligosaccharide

b

5.0

Nutrient (dry matter basis)

Protein

300

Fat

150

Nitrogen-free extract

480

Crude ®ber

15

Calcium

8

Phosphorus

7

Magnesium

1

Sodium

4

Potassium

7

a

Composition of ¯avor and vitamin-mineral premixes are con®dential information.

b

Amounts added prior to extrusion. No direct measures of oligosaccharide concentrations are available.

J.A. Strickling et al. / Animal Feed Science and Technology 86 (2000) 205±219

207

Feed Control Of®cials, 1994). Chromic oxide was added directly to the diet at 2 g/kg (dry

basis) to serve as an indigestible marker of digesta ¯ow. The daily maintenance ration

was based on weight at the beginning of each period as follows:

MJ=d ˆ …293…body weight…kg††

0:75

†…1:6†

Daily ration g=d

…

† ˆ …MJ=d†=…MJ=g†

The rations were calculated based on 16.7 MJ/g feed, with one exception, and that diet

was at 18.8 MJ/g feed to accommodate the eating habit of that particular dog. Daily

rations were divided into two equal meals and fed at 07.00 and 17.00 h. Any food not

consumed within 45 min was removed from the cage and recorded as orts. Water was

provided on a free choice basis. Diets were weighed into stainless steel bowls for each

feeding. From day 2 through day 21, grab samples of each diet were collected and pooled

for diet analysis.

The dogs were allotted randomly to treatments using a 4 7 incomplete Latin square

design structure. The four treatments were based on 5 g/kg of each oligosaccharide

product at the expense of cornstarch as follows: control, fructooligosaccharide (FOS),

mannanoligosaccharide (MOS), and xylooligosaccharide (XOS).

The fructooligosaccharide product used was Raftifeed

1

P75 (Encore Technologies,

Eden Prairie, MN), a powder prepared from the chicory root that contains 0.75

oligofructose. In addition, this product also contains 0.15 fructose, glucose, and

sucrose.

The mannanoligosaccharide product was DP607 (Alltech, Inc., Nicholasville, KY).

These spray-dried mannanoligosaccharides are derived from the yeast cell wall of

Saccharomyces cerevisiae. They are harvested by centrifugation from a lysed yeast

culture.

The xylooligosaccharide product Xylo-oligo 95P (Suntory Limited, Consumer Health

Products Department, Pharmaceutical Division, Tokyo, Japan), used in this experiment

was composed of at least 0.95 xylooligosaccharide in the solid form. The main

components of this xylooligosaccharide product are xylobiose and xylotriose, which are

dimers and trimers of xylose, respectively.

2.3. Sampling

Each experimental period was 21 days. Adaptation diets, half of the diet to be fed for

the new period plus half of the diet fed the previous period, were fed the ®rst day of each

period. The remaining 20 days were at 100% of the experimental diet.

Jugular blood samples were collected on day 7 to test the glucose tolerance of the dogs

to their respective diets. The neck area was shaved the evening prior to the sampling day.

Beginning at 06.00 h catheters (18.5 gauge 5 cm, needle 17 gauge 5 cm; Char-

terMed, Inc., Lakewood, NJ), a 50 cm extension (Baxter Healthcare Corp., Deer®eld, IL),

and a three way stopcock were inserted into the jugular vein of each dog. After all

catheters were inserted, a blood sample was taken (ÿ10 min) to establish the amount of

glucose in the bloodstream prior to diet consumption. In order to ensure that the entire

meal would be consumed within 5 min, half of the normal morning ration was fed. As

208

J.A. Strickling et al. / Animal Feed Science and Technology 86 (2000) 205±219

soon as the meal was complete, another blood sample was taken (0 min) and the time

recorded. Additional samples were taken at 30, 45, 60, 90, 120, 180, 240, 300, and

360 min post 0 min. The ®rst 3 ml of heparinized saline and blood removed was

discarded as waste, after which 5 ml blood samples were taken. Catheters were then

¯ushed with sterile heparinized saline (0.9% NaCl; 20 U/ml heparin, 0.1% benzyl

alcohol). All syringes were heparinized by rinsing with heparin solution (1000 U/ml,

Elkins-Sinn, Inc., Cherry Hill, NJ) prior to sample collection. All blood samples were

immediately stored on ice (maximum 2 h) until they could be brought to the lab,

centrifuged (15 min at 6913 g), plasma removed, and frozen (ÿ208C) for later analysis.

For any dog that removed its catheter during the sampling period, blood was collected

with an 18 gauge needle and syringe for the rest of the day. At the end of the sampling

day catheters were removed.

Fecal collections began at 07.00 h on day 9, ending at 07.00 h on day 13 (5 days). All

feces were removed from the cages and placed in plastic bags, keeping each dog separate.

In addition, dogs were watched carefully during exercise to ensure that feces were

correctly allocated to the proper animal. All feces collected were individually weighed

for each dog and immediately frozen.

Ileal collections began on day 15 and ended on day 17. Dogs were fed at 07.00 h,

cages were cleaned, but the dogs were not exercised. Prior to collection each dog was

®tted with an Elizabethan collar to prevent them from removing the collection bags.

Sterile, 1 oz Whirl-pak

1

bags (Nasco, Fort Atkinson, WI) were placed on each dog

at 08.00, 10.00, 12.00, 14.00, 16.00 h, on day 15 and 16, and at 09.00, 11.00, 13.00,

and 15.00 h on day 17. The bags were removed 1 h later. The dogs were watched

while wearing the collection bags. When a bag became full or started to leak it was

removed and another bag attached. The pH of each collection bag was taken using

an Accumet Basic pH meter (Fisher Scienti®c, Pittsburgh, PA) and recorded for each

dog. Individual ileal samples were weighed, then pooled by dog, and immediately

frozen.

The last 4 days of each period were used to analyze bacteria found in the ileal digesta

and feces. Beginning at 06.00 h on collection days, polyethylene 100 15 mm BioPro

1

petri dishes (International Bioproducts, Inc., Redmond, WA) were labeled in triplicate.

Previously prepared media was melted in an autoclave and placed in a waterbath to

maintain the temperature at 478F. When lab preparation was complete, ileal and fecal

samples were collected from the dogs. The dog's chosen to sample on a given day were

those that voided fresh feces following the morning feeding. Feces in cages upon arrival

each morning were not used. Samples were collected into either ziplock bags (fecal) or

sterile 1 oz Whirl-pak

1

bags (ileal) and immediately brought back to the lab and diluted.

Samples were prepared within 2 h of collection. One dog was not sampled because she

had been on antibiotics at the onset of this experiment for treatment of an infection.

Antibiotics were discontinued prior to the start of any sample collection; however, to

avoid any lingering effects on microorganisms these measurements were not made. It was

felt that the previous antibiotic treatment would have no in¯uence on the digestibility

measures and these were included.

Fecal consistency scores were recorded every day throughout the experiment as

follows.

J.A. Strickling et al. / Animal Feed Science and Technology 86 (2000) 205±219

209

Grade 1 Ð more than two-thirds of the feces in a defecation is liquid. The feces have

lost all form, appearing as a puddle or squirt.

Grade 2 Ð soft-liquid feces; an intermediate between soft and liquid feces.

Approximately equal amounts of feces in a defecation are soft and liquid.

Grade 3 Ð more than two-thirds of the feces in a defecation is soft. The feces retain

enough form to pile but have lost their cylindrical appearance.

Grade 4 Ð firm-soft feces; an intermediate between the grades of firm and soft.

Approximately equal amounts of feces in a defecation are firm and soft.

Grade 5 Ð more than two-thirds of the feces in a defecation are firm. They have a

cylindrical shape with little flattening.
The rating was subjective and performed by the three different people who fed the dogs

during the experiment.

Diet palatability was rated and recorded on day 2 through day 21 of each period.

Grading was divided into four categories as follows.
1. Ate entire meal without hesitation.

2. Ate a portion of the meal but was more concerned with the activity outside of their

cage than the meal itself, but, total meal completed within 1 h.

3. Totally unconcerned about the meal, ate a portion, but, did not consume the entire

ration.

4. Would not or refused to eat.

The rating was subjective and performed by the three different people who fed the dogs

during the experiment.

2.4. Analyses

Both ileal and fecal samples were later thawed and thoroughly mixed so that each

sample was representative of the period. Fresh portions of ileal digesta which were

required for VFA (1 g) and ammonia (1 g) analysis, and fresh portions of fecal samples

(1 g) used for ammonia analysis, were removed. The remainder of the samples were then

weighed into a pan and lyophilized. Weights post-lyophilization were recorded and the

DM coef®cient was calculated. Lyophilized ileal and fecal samples were then ground

with a mortar and pestle until a ®ne, uniform consistency was obtained. Samples were

then dried, in duplicate, in a 708C vacuum oven overnight to a constant weight for DM

determination. Samples were then stored in sealed plastic bags at room temperature.

The diet samples were ®rst thoroughly mixed in a large bowl. Representative samples

of each diet for the four periods were dried in a 558C oven for 46 h for DM determination.

The diet samples were then ground through a 1 mm screen in a Wiley Mill. Dry matters

of ground samples were obtained at 708C as described above, using a 1 g sample of each

diet.

Samples were ashed at 5008C for 16 h and then prepared for Cr analysis as described

previously (Williams et al., 1962) and stored in amber bottles. Chromium analysis was

performed by atomic absorption spectroscopy (Unicam 929 Spectrometer, Thermo Jarell

Ash, Franklin, MA).

210

J.A. Strickling et al. / Animal Feed Science and Technology 86 (2000) 205±219

Previously pooled, fresh, ileal samples were prepared for VFA analysis by mixing 1 g of

each sample with 1 ml of 25% (w/v) metaphosphoric acid and centrifuging (16; 000 g) for

10 min. The supernatant of each sample was removed and analyzed on a Hewlett Packard

5890 Gas Chromatograph (Avondale, PA) with a 1.8 m 4 mm glass column packed

with 10% sp-1000/1% H

3

PO

4

on 100/120 Chromsorb W AW (Supelco, Bellefonte, PA).

For ammonia analysis, the ileal samples previously mixed with metaphosphoric acid for

the VFA analysis, were further diluted with nanopure water to attain a 1:75 dilution. Fresh

pooled fecal samples (0.0125 g) were extracted with 1 ml 0.12N HCl, centrifuged

(13; 000 g) for 4 min, and the supernatant removed. Ammonia was analyzed using

glutamate dehydrogenase (Ammonia Analysis Enzymatic Kit (#171-B), Sigma, St. Louis,

MO) on a Cobas Fara II (Roche Diagnostic Systems, Branchburg, NJ).

Fecal and ileal samples (11 g) were added to 99 ml 0.1% peptone (Difco Labs, Detroit,

MI) diluent (1 g Bacto

1

peptone/l distilled water) and mixed in a blender (10

ÿ1

).

Dilution sequences representing from 10

ÿ2

through 10

ÿ7

g/ml were made using 1 ml of

the previous dilution mixed with 9 ml of peptone diluent. Samples (0.1 ml) were

measured into plates, media was added, plates were stirred for several minutes to assure

that the sample was blended thoroughly into the media, then set aside to gel. The plates

were placed in either the aerobic incubator or the anaerobic chamber, depending on which

bacteria were to be grown. The speci®c bacteria cultured were; C. perfringens (Oxoid

Agar, including supplement A, SR76 and supplement B, SR77, Unipath Ltd.,

Basingstoke, Hampshire, England), Bi®dobacteria (Bacto

1

Liver Veal Agar, Difco Labs,

Detroit, MI) as described previously (McCann et al., 1996), E. coli and total coliforms

(Bacto

1

Violet Red Bile Agar with Mug, Difco Labs, Detroit, MI), Lactobacilli (Bacto

1

Rogosa SL Agar, Difco Labs, Detroit, MI), and anaerobes (Bacto

1

Reinforced Clostridial

Agar, Difco Labs, Detroit, MI). Plates for E. coli and total coliforms were read after a

24 h incubation period using an ultraviolet transilluminator (Fotodyne, Inc., New Berlin,

WI), while the remaining plates were read after a 48 h incubation period. All plates were

digitally counted by the same person to avoid variation in the counting procedure. Dry

matter analysis was performed on both the ileal and fecal samples from the ®rst dilution

(10

ÿ1

).

Jugular plasma samples were thawed, transferred to sample cups and analyzed for

glucose using the Cobas Fara II (Roche Diagnostic Systems, Branchburg, NJ). Plasma

glucose concentrations were determined by enzymatic kit (Glucose (HK) 20, Sigma, St.

Louis, MO).

2.5. Calculation and statistics

Ileal nutrient digestibilities were calculated as described previously (Merchen, 1988)

using Cr as a marker. Each treatment had seven replications except for bacterial samples

which had six. A 4 7 Incomplete Latin Square was the design with dog and period

being the blocking criteria. Analyses were conducted using the General Linear Model

procedure (SAS Inst. Inc, 1988). Orthogonal contrasts were used to separate treatment

means; control versus diets with added oligosaccharide, MOS versus FOS, XOS and FOS

versus XOS. Ileal pH, collected for each ileal sampling, was analyzed using the Mixed

procedure with Satterthwaite's approximation used for the degrees of freedom. The blood

J.A. Strickling et al. / Animal Feed Science and Technology 86 (2000) 205±219

211

plasma collected 11 times post-feeding was also analyzed using the Mixed procedure

with a heterogeneous Compound Symmetry covariance structure for the repeated variable

time, and Satterthwaite's approximation used for the degrees of freedom. Differences

were considered signi®cant with P < 0:05.

3. Results and discussion

3.1. Dry matter

Oligosaccharide addition to diets had no effect on body weight, DM intake, fecal DM

excretion, or DM ¯ow to the ileum (Table 2). Fecal moisture tended to be higher

(P ˆ 0:09) in the control compared with the oligosaccharide diets. Fecal moisture will be

higher in diets containing soluble ®ber (pectin and starch) than insoluble ®ber (corn ®ber

and cellulose; Lewis et al., 1994). However, the differences in ®ber in the present study

were relatively small.

The amount of feces excreted on a DM basis tended to be lower (P ˆ 0:14) in the

control diet than the oligosaccharide diets which was expected. Stool volume has been

shown to increase when oligosaccharides are added to the diet due to the increased

excretion of bacterial biomass (Oku et al., 1984; Tokunaga et al., 1986; Gibson and

Roberfroid, 1995).

Ileal DM ¯ow (g/d) was unaffected by treatment (Table 2). Similar ®ndings have been

reported previously (Gabert et al., 1995) when ileal ¯ow was unaffected by the addition

of 2 g/kg oligosaccharides in weanling pigs. Apparent ileal DM digestibilities also were

not different between diets. Similar results using weanling pigs supplemented with 2 g/kg

oligosaccharides have been reported (Gabert et al., 1995). Oligosaccharides are

Table 2

Evaluation of fructooligosaccharide (FOS), xylooligosaccharide (XOS) and mannanoligosaccharide (MOS)

feeding on dry matter (DM) digestibilities in dogs

Item

Treatments

Contrasts

a

Control FOS

MOS

XOS

S.E.M.

b

Control vs.

others

MOS vs.

FOS, XOS

FOS vs.

XOS

Body weight (kg)

17.3

17.3

17.6

17.7

0.2

0.30

0.70

022

Dry matter intake (g/d)

222

224

222

223

3.0

0.93

0.70

0.88

Fecal moisture (g/d)

350

340

330

350

6

0.09

0.13

0.25

Feces (g DM/d)

33.4

34.9

35.5

36.7

1.3

0.14

0.86

0.33

Ileal ¯ow (g DM/d)

54.2

52.7

52.2

53.2

2.1

0.54

0.78

0.88

DM digestibility coef®cients

Ileal

0.759

0.767

0.772

0.769 0.008

0.29

0.73

0.92

Large intestine

c

0.377

0.330

0.310

0.302 0.034

0.13

0.89

0.57

Total tract

0.852

0.847

0.843

0.839 0.005

0.17

0.95

0.33

a

Probability of greater F-value.

b

Standard error of the mean, n ˆ 7.

c

Coef®cients calculated as fraction of ileal ¯ow.

212

J.A. Strickling et al. / Animal Feed Science and Technology 86 (2000) 205±219

considered to be nondigestible in the small intestine, therefore, it was expected that the

ileal digestibilities in the supplemented diets would be similar to each other. The control

was similar since oligosaccharides were included at only 5 g/kg of the diet.

There were no differences in large intestinal and total tract DM digestibility. Zuo et al.

(1996) fed a low oligosaccharide soybean meal diet to dogs and found it to be more

digestible than a conventional (12 g/kg oligosaccharides; 11 g/kg stachyose and 1 g/kg

raf®nose) soybean meal diet. Although the Zuo et al. (1996) oligosaccharide sources are

different from those used in this experiment, their results indicate that oligosaccharides

can affect digestibility. The 5 g/kg addition of oligosaccharides may not have been

suf®cient to measure effects on DM digestion.

3.2. Nitrogen

There were no differences in N intake or excretion (Table 3). It was expected that

oligosaccharide fermentation in the large intestine would increase fecal N excretion.

Levrat et al. (1993) found that oligosaccharides with relatively high degrees of

polymerization, promoted fecal N excretion in the rat when the level of protein is

moderate. The diet in this experiment was high in protein for the canine and the level of

inclusion was fairly low.

Ileal, large intestinal and total tract N digestibility were unaffected by treatment.

Fructooligosaccharides and XOS supplemented to rats decreased blood urea and renal N

excretion, while there was a corresponding increase in fecal N excretion (Younes et al.,

1995). These effects are dependent on adequate fermentable substrate being presented to

the large intestine for fermentation. The lack of a signi®cant increase in fecal N excretion

in the present study suggests that greater dietary concentrations of oligosaccharides may

be necessary to see such effects in the dog.

Table 3

Evaluation of fructooligosaccharide (FOS), xylooligosaccharide (XOS) and mannanoligosaccharide (MOS)

feeding on nitrogen (N) concentrations in dogs

Treatments

Contrasts

a

Control FOS

MOS

XOS

S.E.M.

b

Control vs.

others

MOS vs.

FOS, XOS

FOS vs.

XOS

N intake (g/d)

10.8

10.8

10.9

11.0

0.1

0.35

0.79

0.30

Fecal N (g/kg)

52

53

52

52

1

0.49

0.50

0.17

Fecal N excreted (g/d)

1.8

1.9

1.9

1.9

0.1

0.16

0.82

0.69

Ileal N (g/kg)

48

48

47

48

1

0.93

0.09

0.92

Ileal N ¯ow (g/d)

2.6

2.6

2.5

2.6

0.1

0.66

0.55

0.91

N digestibility coef®cients

Ileal

0.752

0.758

0.769

0.762

0.010

0.37

0.50

0.83

Large intestine

c

0.319

0.259

0.243

0.254

0.041

0.18

0.79

0.94

Total tract

0.834

0.824

0.825

0.822

0.007

0.20

0.84

0.84

a

Probability of greater F-value.

b

Standard error of the mean, n ˆ 7.

c

Coef®cients calculated as fraction of ileal ¯ow.

J.A. Strickling et al. / Animal Feed Science and Technology 86 (2000) 205±219

213

In contrast, N digestion increased when FOS was incrementally (0±1.5 g FOS/d)

supplemented for 5 days to weanling pigs (Bunce et al., 1995b). Silvio et al. (2000)

studied digestibility in dogs fed diets containing combinations of rapidly (pectin) and

slowly (cellulose) fermented ®bers. When diets contained rapidly fermented ®bers,

fermentation in the large intestine increased, N digestibility in the large intestine

decreased, and total tract N digestibility decreased. The Silvio et al. (2000) study

demonstrates how dramatically ®ber fermentation affects fecal N excretion. Total tract N

digestion decreased from 83 to 73% just by changing source of ®ber.

3.3. Fermentation data

There were no differences in ileal or fecal ammonia concentrations in response to

oligosaccharide feeding (Table 4). This is similar to the ®ndings of Gabert et al. (1995)

who found no effects on ileal ammonia when 2 g/kg oligosaccharides were fed to

weanling pigs and Hussein et al. (1999) who reported no effects on fecal ammonia for

dogs fed diets containing 3, 6 and 9 g/kg oligofructose.

The proportions of acetate, isobutyrate, isovalerate, and valerate in ileal ¯uid were

unaffected by oligosaccharide addition (Table 4). Butyrate proportion tended to be greater

(P ˆ 0:07) for control versus the oligosaccharide supplemented dogs, and lower

(P ˆ 0:09) for the MOS versus the FOS and XOS supplemented dogs. Butyrate

formation is a product of Eubacteria, which are gram-positive, anaerobic bacteria found

as normal inhabitants of the mammalian intestinal tract (Mitsuoka and Kaneuchi, 1977).

Butyrate is also produced when starch is used as a substrate (Wang and Gibson, 1993).

Gabert et al. (1995) reported no differences in cecal VFA concentrations when weanling

pigs were fed 2 g/kg oligosaccharides (galacto, gluco, and lacitol). Howard et al. (1993)

Table 4

Evaluation of fructooligosaccharide (FOS), xylooligosaccharide (XOS) and mannanoligosaccharide (MOS)

feeding on ammonia and ileal volatile fatty acid (VFA) concentrations in dogs

Treatments

Contrasts

a

Control FOS

MOS XOS

S.E.M.

b

Control vs.

others

MOS vs.

FOS, XOS

FOS vs.

XOS

Ileal ammonia (mmol/gDM)

5.1

5.0

4.6

4.4

0.4

0.39

0.87

0.35

Fecal ammonia (mmol/gDM)

25.9

20.9

20.2

25.3

2.0

0.12

0.25

0.14

Ileal VFA (mol/100 mol)

Acetate

68.8

69.2

69.5

69.5

0.5

0.25

0.73

0.68

Propionate

19.1

18.9

19.8

19.1

0.3

0.70

0.09

0.61

Isobutyrate

2.3

2.4

2.1

2.2

0.1

0.77

0.47

0.39

Butyrate

7.6

7.3

6.5

7.0

0.3

0.07

0.07

0.50

Isovalerate

1.5

1.4

1.3

1.3

0.1

0.25

0.26

0.44

Valerate

0.8

0.8

0.8

0.8

0.0

0.67

0.44

0.54

Total ileal VFA (mmol/g DM) 597.6

600.8 554.7 556.3

22.3

0.31

0.40

0.18

Total ileal VFA (mmol/g wet)

79.2

79.5

71.8

73.5

3.1

0.24

0.24

0.19

a

Probability of greater F-value.

b

Standard error of the mean, n ˆ 7.

214

J.A. Strickling et al. / Animal Feed Science and Technology 86 (2000) 205±219

also found no difference in cecal VFA concentrations when neonatal pigs were fed 3 g/l

(liquid diet) fructooligosaccharide.

One might expect to see a higher acetate concentration (Mitsuoka and Kaneuchi, 1977;

Wang and Gibson, 1993), an end-product from the break down of oligosaccharides by

bi®dobacteria. However, the diets had no effect on acetate proportion. Mannanoligo-

saccharides also did not affect cecal concentrations of VFA in chicks from 3 to 10 days of

age (Spring et al., 2000).

3.4. Bacterial concentrations

There were no effects from dietary oligosaccharide addition on any of the bacterial

colony forming units (CFU) plated from the ileal digesta (Table 5). It was hypothesized

that differences in bacterial counts from oligosaccharide feeding found in the ileum

would be minimal since this is not the major site of fermentation. However, counts of

bacteria between the ileum and feces were similar.

In fecal samples (Table 5) the growth of C. perfringens tended to be lower (P ˆ 0:09)

for MOS versus FOS and XOS. Feeding of MOS has been shown to reduce the

colonization of Salmonella in chicks (Spring et al., 2000), but comparable reports for

Clostridial species are lacking.

Literature suggests that an increase in bi®dobacterial counts should be expected with

the addition of fructooligosaccharide or xylooligosaccharide (Wang and Gibson, 1993;

Okazaki et al., 1990). However, others (Mitsuoka and Kaneuchi, 1977) have isolated

Table 5

Evaluation of fructooligosaccharide (FOS), xylooligosaccharide (XOS) and mannanoligosaccharide (MOS)

feeding on concentration of speci®c bacteria (log CFU/g DM)

a

Concentration (log CFU/g DM)

Contrasts

b

Control FOS

MOS

XOS

S.E.M.

c

Control vs.

others

MOS vs.

FOS, XOS

FOS vs.

XOS

Ileal

C. perfringens

4.80

5.21

5.28

5.25

0.4

0.36

0.91

0.94

Bi®dobacteria

10.44

10.60

10.56

10.73

0.1

0.15

0.42

0.44

Lactobacilli

9.37

9.46

10.11

9.60

0.4

0.50

0.31

0.84

Aerotolerant anaerobes

10.50

10.43

10.49

10.36

0.2

0.75

0.68

0.80

E. coli

6.27

6.22

6.42

6.19

0.5

1.00

0.71

0.96

Coliforms

6.70

6.61

7.15

6.86

0.5

0.80

0.55

0.76

Fecal

C. perfringens

4.73

4.74

4.48

5.16

0.2

0.80

0.07

0.16

Bi®dobacteria

10.87

11.05

10.79

10.87

0.1

0.76

0.19

0.23

Lactobacilli

9.32

9.80

10.34

10.15

0.5

0.17

0.54

0.62

Aerotolerant anaerobes

10.59

10.71

10.63

10.61

0.1

0.71

0.88

0.61

E. coli

5.96

5.94

5.80

5.97

0.3

0.83

0.64

0.93

Coliforms

6.28

6.46

6.49

6.61

0.4

0.60

0.92

0.79

a

CFU: colony forming unit.

b

Probability of greater F-value.

c

Standard error of the mean, n ˆ 6.

J.A. Strickling et al. / Animal Feed Science and Technology 86 (2000) 205±219

215

various species of bi®dobacteria and found only three species in the canine as follows: (1)

Bi®dobacteria longum (4.9% of strains isolated), (2) B. adolescentis (70.7% of strains

isolated), and (3) B. pseudolongum (24.4% of strains isolated). When Wang and Gibson

(1993) compared bi®dobacteria growth using oligofructose as a substrate, they found that

all species grew well with the highest speci®c growth rates recorded with B.

pseudolongum, B. infantis, and B. catenulatum, and the lowest growth rate with B.

adolescentis. Two of the highest growth rate species are not found in the canine, and B.

pseudolongum strains, fast growers and high fermenters (Desjardins and Roy, 1990),

comprise only 24.4% of the total bi®dobacteria strains isolated in the canine. In addition,

the lowest growing species, B. adolescentis, comprises 70.7% of the total strains of

bi®dobacteria found in the dog. Therefore, one would expect an increase in the amount of

bi®dobacteria when oligofructose is added to the diet, but not as dramatic as that found in

other experimental animals. Also, Roberfroid and Delzenne (1998) summarized available

literature for humans and reported that although Bi®dobacteria increase their numbers in

response to oligosaccharide feeding, their numbers rarely exceed 10

9.5

, values

comparable to ours when expressed on an `as is' basis (data not shown).

Diet composition is probably the single most important control factor for microbial

activity in the gastrointestinal tract of non-ruminant animals. Digesta reaching the large

intestine determines the fate of the microbial population via the amount that arrives and

the type of substrate that it provides. Diets fed in the present study consisted of 150 g

soyabean meal/kg. The soyabean itself contains a-galactooligosaccharides which pass

along to the large intestine in an intact form where they are fermented by bacteria. These

oligosaccharides may also be used as a selective substrate for the growth of bacteria in the

large intestine. It is possible that the supplemented oligosaccharides were, therefore,

under an additive or masking effect, and that true results of the speci®c oligosaccharides

alone can only be seen if soya products are not used.

Yazawa et al. (1978) reported that B. infantis readily utilized raf®nose and stachyose as

a substrate. However, this bi®dobacteria species is not speci®cally found in the canine and

information was not available on other species. If other bi®dobacteria acted in a similar

fashion to B. infantis, there was approximately 10 g/kg (conventional soyabean meal diet

used by Zuo et al. (1996) contained 185 g soyabean meal, 11 g stachyose, and 1 g

raf®nose/kg) of naturally occurring oligosaccharides included in the base diet which

could have affected the control populations. The theory of a dilution, additive, or masking

effect is even more profound if the log

10

CFU of bi®dobacteria and lactobacilli results

from this experiment are compared to the counts normally found in the canine. According

to Mitsuoka and Kaneuchi (1977) lactobacilli CFU (9:3 1:3) are greater than

bi®dobacteria CFU (6:6 2:7). In both the ileal and the fecal samples from this

experiment, the bi®dobacteria CFU were greater than the lactobacilli CFU.

3.5. Blood glucose

Postprandial plasma glucose concentrations decreased immediately postfeeding

then increased (time effect P ˆ 0:004); however, there was no time by treatment

or treatment effects (data not shown). Glucose concentrations ranged from 87.6 to

100.1 mg/dl.

216

J.A. Strickling et al. / Animal Feed Science and Technology 86 (2000) 205±219

Fructooligosaccharides have been shown to reduce postprandial glycemia and

insulinemia in rats (Roberfroid and Delzenne, 1998). The exact mechanism(s) of this

effect are not understood; however, it is thought to be mediated by decreased gastric

emptying, or changes in hepatic metabolism. Whether these differences in response are

species related (rat versus dog) or dose dependent (100 versus 5 g/kg) are not known.

3.6. Ileal pH

There was no treatment effect on ileal pH. The overall mean ileal pH was 7.1. Similar

results have also been found in the supplementation of oligosaccharides (2 g/kg), for 6

days to weanling pigs that did not alter ileal pH over a 3-day collection (Gabert et al.,

1995). Similarly, neonatal pigs supplemented with 3 g/l of FOS for 15 days resulted in no

change in cecal pH (Howard et al., 1993).

3.7. Fecal consistency

There was no effect of treatment on fecal consistency or palatability scores. The fecal

consistency scores ranged from 4.9 in the FOS and MOS diets to 5.0 in the CON and

XOS diets (data not shown). Oligosaccharides are readily fermentable in the colon.

Supplementation in large quantities could cause acidic fermentation leading to extensive

gas, cramping, and diarrhea. There were no side effects observed indicating that in the

canine, supplementation of oligosaccharides at 5 g/kg is not excessive.

4. Conclusion

Oligosaccharide action is dose dependent as well as structure speci®c. Their addition to

diets is advantageous when they alter the colonic microbiota in favor of bene®cial

bacteria or increase the production of VFA which can be used by the host for energy or for

maintaining healthier gastrointestinal tissues. The overall effect is manipulation of the

ecology of the gastrointestinal tract.

Supplementing the canine with 5 g/kg oligosaccharides produced minor changes in

fecal moisture, ammonia, some ileal VFA and bacteria when compared with a control diet

typical of many commercial feeds. Based on these results, inclusion of 5 g/kg

oligosaccharides in diets similar to the one used in the present experiment, offers little

bene®t for dogs. Whether higher inclusion levels, or supplementation in diets based on

animal protein is bene®cial, remains to be determined.

References

American Association of Feed Control Of®cials, 1994. Petfood Regulations. AAFCO, Atlanta, GA.

Bunce, T.J., Howard, M.D., Allee, G.L., Pace, L.W., 1995a. Protective effect of fructooligosaccharide (FOS) in

prevention of mortality and morbidity from infectious E. coli K:88 challenge. J. Anim. Sci. 73 (Suppl. 1), 69.

Bunce, T.J., Kerley, M.S., Allee, G.L., Day, B.N., 1995b. Feeding fructooligosaccharide to the weaned pig

improves nitrogen metabolism and reduces odor metabolite excretion. J. Anim. Sci. 73 (Suppl 1), 70.

J.A. Strickling et al. / Animal Feed Science and Technology 86 (2000) 205±219

217

Cummings, J.H., Macfarlane, G.T., 1991. The control and consequences of bacterial fermentation in the human

colon. J. Appl. Bacteriol. 70, 443±459.

Delzenne, N., Kok, N., Kok, N., Deboyser, D., Goethals, F., Roberfroid, M., 1993. Dietary fructooligosacchar-

ides modify lipid metabolism in the rat. Am J. Clin. Nutr. 57, 820S.

Desjardins, M., Roy, D., 1990. Growth of bi®dobacteria and their enzyme pro®les. J. Dairy Sci. 73, 299±

307.

Gabert, V.M., Sauer, W.C., Mosenthin, R., Schmitz, M., Ahrens, F., 1995. The effect of oligosaccharides and

lacitol on the ileal digestibilities of amino acids, monosaccharides and bacterial populations and metabolites

in the small intestine of weanling pigs, monosaccharides and bacterial populations and metabolites in the

small intestine of weanling pigs. Can. J. Anim. Sci. 75, 99±107.

Gibson, G.R., Roberfroid, M.B., 1995. Dietary modulation of the human colonic microbiota: introducing the

concept of prebiotics. J. Nutr. 125, 1401±1412.

Hidaka, H., Eida, T., Takizawa, T., Tokunaga, T., Tashiro, Y., 1986. Effects of fructooligosaccharides on

intestinal ¯ora and human health. Bi®dobacteria Micro¯ora 5, 37±50.

Howard, M.D., Kerley, M.S., Gordon, D.T., Pace, L.W., Garleb, K.A., 1993. Effect of dietary addition of

fructooligosaccharide on colonic micro¯ora populations and epithelial cell proliferation in neonatal pigs.

J. Anim. Sci. 71 (Suppl 1), 177.

Hussein, H.S., Flickinger, E.A., Fahey Jr., G.C., 1999. Petfood applications of inulin and oligofructose. J. Nutr.

129, 1454S±1456S.

Levrat, M., ReÂmeÂsy, C., DemigneÂ, C., 1993. In¯uence of inulin on urea and ammonia in the rat cecum:

consequences on nitrogen excretion. J. Nutr. Biochem. 4, 351±356.

Lewis, L.D., Magerkurth, J.H., Roudebush, P., Morris Jr., M.L., Mitchell, E.M., Teeter, S.M., 1994. Stool

characteristics, gastrointestinal transit time and nutrient digestibility in dogs fed different ®ber sources.

J. Nutr. 124, 2716S±2718S.

McCann, T., Egan, T., Weber, G.H., 1996. Assay procedures for commercial probiotic cultures. J. Food

Protection 59, 44±45.

Merchen, N.R., 1988. Digestion, absorption and excretion in ruminants. In: Church, D.C. (Ed.), The Ruminant

Animal: Digestive Physiology and Nutrition. Prentice Hall, Englewood Cliffs, NJ, pp. 188±189.

Mitsuoka, T., Kaneuchi, C., 1977. Ecology of the bi®dobacteria. Am. J. Clin. Nutr. 30, 1799±1810.

Newman, K., 1994. Mannan-oligosaccharides: natural polymers with signi®cant impact on the gastrointestinal

micro¯ora and the immune system. In: Lyons, T.P., Jaques, K. A. (Eds.), Biotechnology in the Feed Industry.

Proceedings of Alltech's Tenth Annual Symposium. Nottingham University Press, Nottingham, UK, pp. 167±

174.

Okazaki, M., Fujikawa, S., Matumoto, N., 1990. Effect of xylooligosaccharide on the growth of bi®dobacteria.

Bi®dobacteria Micro¯ora 9, 77±86.

Oku, T., Tokunaga, T., Hosoya, N., 1984. Nondigestibility of a new sweetener, Neosugar, in the rat. J. Nutr. 114,

1574±1581.

Roberfroid, M., Delzenne, N., 1998. Dietary fructans. Annu. Rev. Nutr. 18, 117±143.

SAS Inst. Inc., 1988. SAS Language Guide for Personal Computers, Release 6.03. Cary, NC.

Silvio, J., Harmon, D.L., Gross, K.L., McLeod, K.R., 2000. In¯uence of ®ber fermentability on nutrient

digestion in the dog. Nutrition 16, 289±295.

Spring, P., Wenk, C., Dawson, K.A., Newman, K., 2000. The effects of dietary mannanoligosaccharides on cecal

parameters and the concentrations of enteric bacteria in the ceca of salmonella-challenged broiler chicks.

Poultry Sci. 79, 205±211.

Stephen, A.M., Cummings, J.H., 1980. The microbial contribution to human fecal mass. J. Med. Microbiol.

45±56.

Tokunaga, T., Oku, T., Hosoya, N., 1986. In¯uence of chronic intake of new sweetener, Neosugar

(fructooligosaccharide), on growth and gastrointestinal action of the rat. J. Nutr. Sci. Vitaminol. (Tokyo)

32, 111±121.

Walker, J.A., Harmon, D.L., Gross, K.L., Collings, G.F., 1994. Evaluation of nutrient utilization in the canine

using the ileal cannulation technique. J. Nutr. 124, S2672±S2676.

Wang, X., Gibson, G.R., 1993. Effects of the in vitro fermentation of oligofructose and inulin by bacteria

growing in the human large intestine. J. Appl. Bacteriol. 75, 373±380.

218

J.A. Strickling et al. / Animal Feed Science and Technology 86 (2000) 205±219

Williams, C.H., David, D.J., Iismaa, O., 1962. The determination of chromic oxide in faeces samples by atomic

absorption spectrophotometry. J. Agric. Sci. 59, 381±385.

Yamashita, K., Kawai, K., Itakura, M., 1984. Effects of fructooligosaccharides on blood glucose and serum

lipids in diabetic subjects. Nutr. Res. 4, 961±966.

Yazawa, K., Imai, K., Tamura, Z., 1978. Oligosaccharides and polysaccharides speci®cally utilizable by

bi®dobacteria. Chem. Pharm. Bull. 26, 3306±3311.

Younes, H., Garleb, K.A., Behr, S., ReÂmeÂsy, C., DemigneÂ, C., 1995. Fermentable ®bers or oligosaccharides

reduce urinary nitrogen excretion by increasing urea disposal in the rat cecum. J. Nutr. 125, 1010±1016.

Zuo, Y., Fahey Jr., G.C., Merchen, N.R., Bajjalieh, N.L., 1996. Digestion responses to low oligosaccharide

soybean meal by ileal-cannulated dogs. J. Anim. Sci. 74, 2441±2449.

J.A. Strickling et al. / Animal Feed Science and Technology 86 (2000) 205±219

219