| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ANIMAL NUTRITION |
Animal Nutrition Group, Wageningen University and Research Centre, PO Box 338, 6700 AH, Wageningen, the Netherlands
 |
Abstract |
|---|
 Top Abstract INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: fermentable carbohydrate • fermentation • weaning piglet
|
INTRODUCTION |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). There may be possibilities to improve pig health by stimulating the commensal microbiota.
The gut microbiota is dependent upon the animal diet as the main source of substrate for its metabolism. Any changes in diet composition may have effects on the intestinal microbiota (Bedford and Apajalahti, 2001). Depletion in carbohydrates as energy source leads to more proteolytic fermentation in the hindgut (Piva et al., 1996). Excess protein fermentation in the large intestine results in increased ammonia concentration in the colon and predisposes early-weaned piglets to diarrhea (Dong et al., 1996). Sometimes temporary post-weaning anorexia is followed by a sudden increase in feed intake, which leads to an upset of the digestive function and to diarrhea (Hopwood and Hampson, 2003).
Inclusion of fermentable carbohydrates in the diet is an effective strategy to control microbial proteolysis (Shi and Noblet, 1993; Piva et al., 1996; Houdijk Jos, 1998). In pigs, substantial fermentation has been observed in the small intestine (Drochner, 1991; Jensen and Jorgensen, 1994; Konstantinov et al., 2004). By selecting ingredients with different rates of fermentation, it should be possible to design a diet that will stimulate fermentation all along the GIT.
The current study aimed to investigate 1) whether the inclusion of fermentable carbohydrates with variable rates of fermentation in weaning diets influences fermentation along the GIT and reduces the protein fermentation, 2) whether enforced fasting at the beginning of the postweaning period has an effect on concentration of fermentation end products in later periods, and 3) whether there are any gradual changes in the fermentation end-product profiles with time after weaning.
|
MATERIALS AND METHODS |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Experimental Design
An in vivo experiment was designed as 2 x 2 factorial arrangements of treatments. The experimental treatments were the 2 dietary treatments (with and without inclusion of fermentable carbohydrates) and 2 feeding management treatments (with or without enforced fasting at the beginning of the experimental period). The experiment was conducted with 3 identical groups of piglets. Each group was evaluated over a period of 10 d. For each group, 4 piglets from 4 litters were used (16 piglets per group; 48 piglets in total for the experiment). The piglets were chosen randomly.
On d 1 of each period, fecal samples were collected before the piglets were subjected to any treatment. Then, piglets from each litter were assigned to 1 of the 4 treatment combinations. Daily fecal samples were subsequently collected. At the end of each period, the piglets were slaughtered and digesta samples were collected from different parts of the GIT.
Animals and Housing
The 48 crossbred piglets (Hypor x Pietrain, mixed group of males and females) were weaned at 4 wk of age and transported to the experimental facility. The piglets had no access to creep feed before weaning, nor any antibiotic treatment before or during the experimental period. During the experimental period of 10 d, the piglets had free access to their diet (except the fasted piglets for the first 48 h) and clean drinking water. Daily feed intake was measured during the experiment. Piglets from the same litter were kept in adjacent individual pens separated by a wire mesh, so that they could have visual contact with their littermates. This arrangement was to prevent cross-contamination between litters. The continued contact with littermates was done to potentially reduce stress.
Dietary Treatments
The control diet (CON) was semipurified and was designed to have very low levels of fermentable carbohydrates (Table 1). The test diet with added fermentable carbohydrates (CHO) was based on this same diet, but had added carbohydrates in the form of unmolassed sugar beet pulp, native wheat starch, lactulose, and inulin. In the current study, 4 fermentable carbohydrates with different rates of fermentation were included in 1 of the dietary treatments.
|
). Lactulose and inulin were more rapidly fermentable compared with wheat starch and sugar beet pulp. Levels of these carbohydrates were chosen to keep the dietary total energy of both diets comparable. Within fermentable carbohydrates in the CHO diet, priority was given to sugar beet pulp and wheat starch, considering they would reach the hindgut where the amount and diversity of microbiota is greater compared with the small intestine, where inulin and lactulose are most likely to be fermented.
For both of the diets, the main source of starch used was native cornstarch, with an ileal digestibility of approximately 97% (Martinez-Puig et al., 2003), to have a contrast in the substrate reaching the large intestine. The diets were composed in such a way that total energy and protein contents were comparable. Both diets contained no antibiotics and no copper beyond that of the trace mineral premix. The composition of the diets is shown in Table 1
.
Fasting Treatments
The animals with enforced fasting were not offered any feed for 48 h from the moment of arrival at the experimental facility. The nonfasted animals, on the other hand, had free access to their diet from the moment of arrival at the facility. All piglets had free access to water at all times.
Slaughtering and Sampling
Daily fecal samples were collected in the morning with a gloved finger for DM, VFA, and ammonia concentration analyses. The piglets were slaughtered on d 10. First, Ketamin (Sanaket 10%, Anisane B.V., Raamsdonksveer, the Netherlands) was used as a preanesthetic (15 mg/kg of BW); 30 min later, the piglet was euthanized by intracardiac injection of T61 (a combination of embutramide, mebenzoniumiodide, and tetracain hydrochloride; Hoechst Roussel Vet, Frankfurt, Germany).
After dissection into the abdomen, the entire GIT was separated from the abdominal cavity. The GIT was ligated at regular intervals to avoid mixing of the digesta. In the laboratory, the GIT was divided into 4 parts: the first (cranial) half of the small intestine (SI-1), the second half of the small intestine (SI-2), the cecum (CE), and the colon (CO). From each part, digesta were mixed, and samples were collected for DM, VFA, and ammonia analysis within 10 min of slaughter. Samples were stored at –20°C until the analyses were done.
Analyses
Dry matter was determined by drying to a constant weight at 103°C (ISO standard 6496; ISO, 1999) and ash by combustion at 550°C (ISO standard 5984; ISO, 1978). The VFA concentrations in the fermentation liquids were analyzed by gas chromatography (Fisons HRGC Mega 2, CE Instruments, Milan, Italy), using a glass column fitted with Chromosorb 101; N2 saturated with methanoic acid as the carrier gas, at 190°C; and isocaproic acid as an internal standard. Ammonia was determined according to the method described by Searle (1984).
Calculations
The branched-chain proportion (BCP) was calculated as an indicator of protein fermentation (Macfarlane et al., 1992), as follows:
 | [1] |
The acetic (AP), propionic (PP), and butyric (BP) acid molar proportions (%) were calculated similarly.
Statistics
All statistical analyses were performed using the PROC GLM procedure of SAS (SAS Inst., Inc., Cary, NC). Because concentrations of the fermentation end product in feces and daily feed intake of the animal were measured repeatedly during the experiment, the statistical analysis was done for repeated measurements. In short, for the separation of time from treatment effects, a straight line was fitted through the end-product concentrations or daily feed intake values against time, beginning with the measurement from d 2. The existence of a quadratic component in this relationship was tested, but because it failed to improve the statistical fit over the linear, the latter was preferred. The slope of the line (
) represented the effect of time (t), and the intercept (at t = t) represented the average end-product concentration or daily feed intake during the experiment. Both the slope and intercept were treated as dependent variables, and treatment effects on slope and intercept were analyzed by ANOVA (see Eq. 2).
The values from d 1 feces were not considered in the analysis because on d 1, fecal samples were collected before the piglets were exposed to the treatments. The difference between the intercept for different treatment combinations and d 1 values for end-product concentrations was considered the postweaning change. This change was analyzed for significant difference among treatment combinations by ANOVA, using the following model:
 | [2] |
where Y is the parameter to be tested, µ is the overall mean, Di represents the effect of the diet i (i = 1, 2), Fj represents the effect of the fasting treatment j (j = 1, 2), (D x F)ij represents the interaction of diet and fasting, and 
ijk is the error term. The effect of period and litter was tested separately, and was removed from the model because it had no effect on any of the parameters tested.
For the digesta samples on d 10, the effects of diet, fasting, and the site within the GIT (where feces on d 10 were also considered as an additional site), and the interaction between diet and site within GIT, were analyzed by ANOVA, as follows:
 | [3] |
where Y was the parameter to be tested; µ represented the overall mean; Di represented effect of the diet I; Fj represented the effect of the fasting treatment; (D x F)ij represented the effect of the interaction of diet and fasting; 1ijk was the error term 1, which represented the random effect of animal within diet i, and fasting treatment j; Gl represented the effect of site within the GIT l; (D x G)il, (F x G)jl, and (D x F x G)ijl denoted the respective interactions; and 2ijklm was the error term 2, which represented the overall error. Differences were considered significant when P < 0.05.
|
RESULTS |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
. Most of the responses, including ADFI and fecal DM content, changed with time (P < 0.001) except ammonia and total VFA concentration. This effect of time was independent of the experimental treatments with the exception that the effect of time on ADFI was affected by fasting (P = 0.018).
|
. Fasting had no effect on any of the responses nor was there any interaction with diet. Diet had an effect on the ammonia concentration (P = 0.03) and BCP (P = 0.04), which were both lower for the CHO diet compared with the CON diet. Diet had an effect on the DM content of the feces (P = 0.05) but no effect on the total VFA concentrations.
|
. For both dietary groups, there was a postweaning increase in the total VFA concentrations (for CON diet, P = 0.019 and for CHO diet, P = 0.015). However, ammonia concentration (P < 0.001 for both diets), AP (P = 0.007 for CON diet and P = 0.017 for CHO diet), and BCP (P = 0.002 for CON diet and P < 0.001 for CHO diet) were decreased postweaning. The diets had an effect on the postweaning change in terms of BCP and ammonia concentration. The CHO diet was observed to be more effective in lowering BCP (P = 0.04) and ammonia (P < 0.01) concentrations compared with the CON diet, but failed to increase total VFA concentration.
|
). Although the total VFA concentrations were greater for the CHO group (Awati, 2005), the difference between dietary groups did not reach significance. The GIT segment had effects (P < 0.001) on all responses including DM content, VFA, and ammonia concentrations. However, by d 10, fasting had no effect on the fermentation end products in the different sites within GIT, nor any interaction with diet.
|
|
DISCUSSION |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). Fasting had no effect on any of the average fermentation end-product profiles in feces. This might be the result of low feed intake on d 1 and 2 for piglets from the nonfasted group as well. In the nonfasted group, independent of diet, the average cumulative feed intake for the first 2 d postweaning was 52 g, which was consistent with the value (<100 g) reported by McCracken et al. (1999), but low compared with 110 g observed on the first day by Pluske et al. (1996) from the piglets offered a dry starter diet. In the current study, both fasted and nonfasted groups had comparable feed intake from d 3 onwards (Figure 3). There was no difference between the 2 groups for ADFI (Table 3). The ADFI for both groups was comparable with that (286 g) observed by Pluske et al. (1996). During the experimental periods, none of the piglets showed any signs of diarrhea or any other illness. These observations lead to the conclusion that either 1) the results of the current study do not support the theory of temporary anorexia after weaning followed by a large feed intake that leads to greater protein fermentation (as suggested by Makkink, 1993) or postweaning diarrhea (as observed by Hopwood and Hampson, 2003), or 2) the sudden rise of approximately 200 g in feed intake on d 3 observed in the fasted group was not large enough to constitute overeating and subsequent gastric disturbances.
|
followed this approach for creep feed intake studies.
Time after weaning plays a major role in fermentation end-product profiles in the feces (Table 2). The AP, PP, and BP were time dependent (P < 0.001), although the effects of time on ammonia and total VFA concentration were less clear. The PP and BP were increased and the AP decreased (P < 0.001) with time after weaning. This might be due to postweaning development of microbiota, both in terms of numbers as well as in diversity. Furthermore, with the increase in ADFI with time (Table 2), availability of the substrate for the large intestinal microbiota was also increased. With diverse microbiota and high substrate availability, there is greater PP and BP, whereas AP decreases considerably (Macfarlane and Macfarlane, 2003). The DM content of the feces was increased with time (Table 2). Absorption of short-chain fatty acids stimulates sodium absorption from the intestinal lumen, which adds to the efficient reabsorption of the water in the large intestine (Roediger, 1980). Furthermore, the colon increases in size in the postweaning period (Pluske et al., 2003), which adds to the absorptive capacity of the colon, resulting in increasing DM content of feces.
Although, in comparison with the CON diet, the CHO diet resulted in lower ammonia concentrations (P = 0.03) and BCP (P = 0.04), which are mainly products of protein fermentation (Yokoyama et al., 1982; Russell et al., 1983; Macfarlane et al., 1992), there were no differences for VFA concentrations in the feces due to diet (Table 3). It was expected that cornstarch with approximately 97% ileal digestibility (Martinez-Puig et al., 2003) would reach the large intestine in negligible amounts. But the decrease in amylase activity during the first week after weaning (Lindemann et al., 1986) might have led to the escape of some cornstarch from enzymatic digestion in the small intestine and becoming available to microbial fermentation in the large intestine. This may therefore have masked the contrast in the 2 dietary treatments in the current study.
Postweaning change in fermentation characteristics showed a decrease in ammonia (P < 0.001 for both diets) and BCP (P = 0.002 for CON diet and P < 0.001 for CHO diet). This decrease was facilitated by the CHO diet (Figure 1). In the current study, both diets contained native cornstarch as the main starch source, along with fermentable carbohydrates in the CHO diet (Table 1). This clearly demonstrates a shift from protein fermentation toward carbohydrate fermentation.
In different parts of the GIT, a significant reduction in ammonia (cecum, colon, and feces) and BCP (SI1) on d 10 was also observed with the CHO diet compared with the CON diet (Figure 2). Similar results had been confirmed in an earlier in vivo study in our laboratory (Awati, 2005) for animals fed the same diets. This supports the hypothesis that inclusion of fermentable carbohydrates with different fermentation rates in weaner diets improves carbohydrate fermentation along the GIT with a reduction in harmful protein fermentation.
|
Footnotes |
|---|
2 Corresponding author: a.awati{at}massey.ac.nz
Received for publication December 7, 2004. Accepted for publication March 23, 2006.
|
LITERATURE CITED |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. Pie, A. Awati, S. Vida, I. Falluel, B. A. Williams, and I. P. Oswald Effects of added fermentable carbohydrates in the diet on intestinal proinflammatory cytokine-specific mRNA content in weaning piglets J Anim Sci, March 1, 2007; 85(3): 673 - 683. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
