| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Danish Institute of Agricultural Sciences, Department of Animal Nutrition and Physiology, Research Center Foulum, Tjele, Denmark
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
0.02). Fermented liquid feed caused a decrease in gastric pH from 4.4 and 4.6 for DF and NLF, to 4.0 (P = 0.003), and increased numerically the gastric concentration of lactic acid (P = 0.17) from 50 to 60 mmol/kg in the DF and NFLF treatments to 113 mmol/kg in the FLF treatment. The animals fed NFLF showed the highest weight gain (995 g/d) and feed intake (2.14 kg/d), and those fed FLF the lowest values (weight gain, 931 g/d; feed intake, 1.96 kg/d; P = 0.003 for weight gain, and P < 0.001 for feed intake). The results from the present study indicate that feeding FLF as prepared here may be a valid feeding strategy to decrease the levels of enterobacteria in the GI-tract of growing pigs, whereas feeding liquid feed that has started to ferment (high levels of enterobacteria and high pH as with NFLF) increases the presence of these undesirable bacteria. Nonetheless, higher daily feed intake and body weight gain are obtained when feeding NFLF compared with feeding FLF or DF.
Key Words: Digestive Tract • Fermented Foods • Growth • Liquid Diets • Microbial Flora
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
; Scholten, 2001). Benefits of liquid feeding include its positive effects on the gastrointestinal microflora (e.g., reduction in Salmonella) and its potential to utilize coproducts from the food industry (Jensen and Mikkelsen, 1998; Geary et al., 1999; Scholten et al. 2002).
It is important to make a distinction between nonfermented liquid feed (NFLF) and fermented liquid feed (FLF). The former is defined here as a mixture of feed and water made immediately before feeding or in the trough at feeding, whereas the latter denotes a mixture of feed and water stored in a tank at a certain temperature and for a certain period of time before it is fed to the animals. From the moment feed and water are mixed, there is a possibility that fermentation will begin. The initial phase of fermentation is characterized by low levels of lactic acid bacteria, yeasts, and lactic acid, high pH, and, importantly, a blooming of enterobacteria. This phase is followed by a second phase, in which a steady state is reached, and which is characterized by high levels of lactic acid bacteria, yeasts, and lactic acid, low pH, and low enterobacteria counts (Jensen and Mikkelsen, 1998; Canibe et al., 2001; Lawlor et al., 2002).
Fermented liquid feed has been reported by several authors to decrease the levels of enterobacteria along the gastrointestinal tract (GI-tract) of pigs and piglets compared with dry feed and NFLF (Mikkelsen and Jensen, 1997; Moran, 2001; van Winsen et al., 2001). Feeding FLF decreases Salmonella seroprevalence in pigs (van der Wolf et al., 2001) and incidence of dysentery in pigs infected with Brachyspira hyodysenteriae (Lindecrona et al., 2000) compared with dry feed. On the other hand, feeding liquid compound feed soaked in a trough for a few hours has been shown to increase the risk of Salmonella infection in pigs (van der Wolf et al., 1999).
Data on the effect of feeding NFLF or FLF to growing pigs on growth performance is scarce (reviewed by Jensen and Mikkelsen, 1998; Hurst et al., 2001; Pedersen et al., 2002).
The present study was designed with the main purpose of investigating the characteristics of NFLF and FLF, and their effect on some aspects of the microbial ecology of the GI-tract, that is, microbial population size and metabolism, compared to dry feed when fed to growing pigs.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
Diets and Feeding
A grower diet (unheated) was formulated (Table 1). Three dietary treatments were designed using this diet: 1) dry feed supplied as meal (DF), 2) NFLF, and 3) FLF. The liquid feed was prepared by mixing meal and water in a 1:2.5 ratio. The NFLF was fed immediately after mixing the meal and water. The FLF was prepared by storing the liquid feed in a closed tank under agitation at 20°C for 4 d before being offered to the pigs; at each feeding, 50% of the content was taken out and replaced in the tank with an equal amount of fresh feed and water. The animals were fed following a standard restrictive feeding scale according to their BW (so that the whole ration was consumed within approximately 30 min) twice daily, at 0730 and at 1430. Additional fresh water was available for all pigs from nipple drinkers.
|
Experimental Procedure
A sample of each of the three diets was taken weekly immediately before the morning meal, brought to the laboratory, and analyzed within 2 h. The pH was measured, subsamples for organic acid determination and DM determination were taken and frozen, and another subsample was transferred to a salt medium (as described in Analytical Methods). The period from collection of samples to plating could be more than 2 h, but when the samples are in the salt medium, little to no growth or death of microorganisms is expected.
Body weight and feed intake were recorded every fortnight to a final BW of 101 kg, SD = 4.0. At a BW of 112 ± 6.6 kg, five animals per treatment were sacrificed 3 h after the morning meal with a captive bolt pistol. The GI-tract was immediately removed and divided into eight segments: stomach, three equal segments of the small intestine, the cecum, and three equal segments of the colon, including the rectum. The total content of each segment was weighed, and pH was immediately measured. Immediately after sample collection, approximately 5.0 g of sample was extracted with 10.0 mL of PCA/EDTA (2 M cold perchloric acid containing 10 mM EDTA) and stored at -80°C for ATP and adenylate energy charge (AEC) analyses. Dry matter, short chain fatty acids (SCFA), ATP and AEC determinations were carried out on the contents from all segments. Microbiological determinations were immediately performed on digesta from the stomach, the distal segment of the small intestine, the cecum, and the middle segment of the colon.
Analytical Methods
Dry matter content of diets and digesta was determined by freeze-drying the sample. Furthermore, to express the results of chemical analyses in DM%, DM was determined by drying at 103°C for 20 h. Ash and nitrogen were determined according to AOAC (1990) methods, energy by a Leco AC 300 automated calorimeter system 789-500 (Leco, St. Joseph, MI). Amino acid analyses were carried out according to Mason et al. (1980). Fat was extracted with diethyl ether after acid hydrolysis and analyzed as described by Stoldt (1952). Low molecular weight (LMW) sugars were measured following a modification of the method of Bach Knudsen and Li (1991) as described by Bach Knudsen (1997) and nonstarch polysaccharides (NSP) were measured by a modification of the Uppsala method and that of Englyst et al. (1982), as described by Bach Knudsen (1997). The concentration of SCFA and lactic acid was assayed by the method of Jensen et al. (1995). Ethanol was determined according to Beutler (1984).
Feed (10 ± 3.5 g) and digesta samples (7 ± 1.9 g) were transferred rapidly after collection under a flow of CO2 into flasks containing 90 mL of a prereduced salt medium (Holdeman et al., 1977). The suspension was then transferred to a CO2-flushed plastic bag and homogenized in a stomacher blender (Seward Medical, London, U.K.) for 2 min. Then, 10-fold dilutions were prepared in prereduced salt medium by the technique of Miller and Wolin (1974). Samples (0.1 mL) were plated on or inoculated to both selective and nonselective media. Total anaerobic bacteria were enumerated by culturing the samples in roll tubes containing ruminal fluid-glucose-cellobiose agar (Holdeman et al., 1977) and incubating anaerobically at 37°C for 7 d. Lactic acid bacteria were enumerated on de Man, Rogosa, and Sharp agar (Merck 1.10660, Darmstadt, Germany) following anaerobic incubation at 37°C or 20°C for 2 d. Enterobacteria were enumerated on McConkey agar (Merck 1.05465) following aerobic incubation at 37°C for 1 d. Yeasts were enumerated on malt chloramphenicol agar (10 g/L glucose (Merck 1.08337.1000); 5 g/L of malt extract (Difco 0186-17-7, Detroit, MI); 10 g/L peptone (Difco 0118-17-0); 50 mg/L chloramphenicol (Sigma-Aldrich Chemie GmbH C-0378, Steinheim, Germany); 15 g/L agar (Merck 1.01614) following aerobic incubation at 37°C or 20°C for 2 d.
The rate of SCFA and lactic acid production in the GI-tract of pigs fed the three diets was determined by incubating 10 g of gastrointestinal contents from all eight segments with 40 mL of 100 mM Na-phosphate buffer (pH 6.5). The incubations were essentially carried out as described by Jensen and Jensen (1993), with an incubation time of 2 h in serum bottles. Samples were taken at 0, 1, and 2 h.
The concentrations of ATP and AEC were determined by the luciferin-luciferase method as described by Jensen and Jørgensen (1994).
Calculations
When diet or digesta samples had counts below detectable levels, the minimum detectable level was applied. When the minimum detectable level was applied to one treatment, statistical differences between treatments only indicated minimal differences; when it was applied to two or all three treatments for a given microorganism, the P-value was only an approximation.
The concentration of undissociated lactic acid in the stomach was calculated using the Henderson-Hasselbach equation:
 |
where pH is the pH measured in the stomach, [A-] represents the concentration of dissociated acid, and [HA] represents the concentration of undissociated acid. The pKa value for lactic acid is 3.84. The rate of organic acid production was linear during the 2 h of incubation, so the rate of production was calculated using the slope of the lines between 0 and 2 h. Before statistical analysis of microbial counts, logarithmic conversion of the data was performed.
Statistical Methods
The effect of diet in a given intestinal segment, and the effect on gain:feed ratio were tested using a simple ANOVA based on the following GLM:
 |
where Yij is the dependent variable, µ is the overall mean; 
i is the effect of diet, i = 1, 2, 3; 
j is the effect of litter, j = 1, ..., 5 for data from the GI-tract, j = 1, ..., 20 for data on gain to feed ratio; and 
~ N(0,
2) represents the unexplained random error.
The effect of diet on ADG and ADFI was analyzed with initial weight as covariate, following the GLM:
 |
where Yij is the dependent variable, µ is the overall mean; i is the effect of diet, i = 1, 2, 3; j is the effect of litter, j = 1, ..., 20; IWij is the initial body weight, i = 1, 2, 3, and j = 1,..., 20; and ~ (0,2) represents the unexplained random error.
The analyses were performed with SAS for Windows, version 6.12 (SAS Inst., Inc., Cary, NC). When there was an overall effect of diet, differences between means were compared by Tukey’s least significant difference. A significant difference was declared at P < 0.05.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
. The concentration of lysine (g/16 g of N) was highest in the DF diet (6.0), somewhat lower in the NFLF (5.8) diet, and lowest in the FLF (4.8) diet. The concentration of the other AA was similar for all diets (data not shown). The levels of LMW sugars were reduced by fermentation, with values decreasing from 3.6% in the DF to 2.9% in the NFLF and 0.1% in the FLF. The pH for the NFLF was 5.98 and 4.36 for the FLF (Table 3). In the FLF, of 24 samples, three showed values higher than 4.50 (considered to be the maximal level to obtain FLF of good quality; Jensen and Mikkelsen, 1998). The concentration of SCFA was similar in the DF and NFLF diets and highest in the FLF diet, acetic acid being the main contributor to the difference between diets. Lactic acid was not detected in the DF diet, low levels were measured in the NFLF (approximately 1 mmol/kg), and the highest levels were measured in the FLF (approximately 169 mmol/kg). Of the total lactic acid in the FLF, 90% was L-lactic (N = 5). The microbial counts of the diets (Table 4) showed that NFLF contained higher microbial counts than DF. The FLF contained the highest counts of lactic acid bacteria grown at 20 and 37°C (9.4 log cfu/g), yeasts grown at 20°C (6.9 log cfu/g), and total anaerobes (8.5 log cfu/g), whereas this diet contained the lowest levels of enterobacteria (<3.2 log cfu/g) compared with the other two experimental diets. Ethanol concentration in the FLF (n = 5) diet was 0.14 ± 0.015 g/kg of sample.
|
|
|
). Feeding NFLF resulted in the lowest pH values in the mid- and distal small intestine and distal colon.
|
). The proximal small intestine content of animals fed both NFLF and FLF (13.9 to 14.0%) had the lowest DM content compared to that of animals fed DF (17.2%) (P = 0.02).
|
). Feeding NFLF resulted in higher gastric density of lactic acid bacteria grown at 20°C than when feeding DF (P < 0.001) and in the highest level of enterobacteria in the stomach and small intestine compared with feeding DF or FLF (P < 0.001), with the same tendency in the cecum and mid-colon.
|
) did not differ between diets, except for a higher AEC value in the mid- and distal colon of pigs fed FLF compared with pigs fed the DF or NFLF diet.
|
0.09), the values varying from 63 to 113 mmol/kg for FLF and from 40 to 62 mmol/kg for DF and NFLF (Table 8). Formic, acetic, propionic, butyric, isobutyric, valeric, isovaleric, and capronic acid were measured in all eight segments of the GI-tract. The sum of acetic, propionic, and butyric acid was not significantly different between diets, increasing from the stomach (23 to 36 mmol/kg) and distal small intestine (19 to 27 mmol/kg) to the proximal colon (128 to 143 mmol/kg), after which the values decreased to 103 to 112 mmol/kg in the distal colon. There was an effect of feeding FLF on the ratio of acetic acid to the sum of acetic + propionic + butyric acid (P = 0.001) and propionic acid (P < 0.001) in the stomach. The ratio of acetic acid was 63 to 66% for the DF and NFLF diets compared with 84% for FLF, and the values for propionic acid were 22 to 23% for DF and NFLF diets, compared with 10% for FLF. The ratio of butyric acid in the stomach was numerically lower for FLF (6%) compared with DF and NFLF (12 to14%) (P = 0.15). Isobutyric acid and isovaleric acid were not detected in the proximal GI-tract, and in the cecum and colon, the values were similar for all diets. The values increased from 1.2 to 1.8 mmol/kg in the cecum to 4.0 to 5.5 mmol/kg in the distal colon. The concentration of undissociated lactic acid in the stomach was 10.9 mmol/kg (SEM = 4.97), 6.8 mmol/kg (SEM = 5.82), and 45.5 mmol/kg (SEM = 4.97) for the pigs fed DF, NFLF, and FLF, respectively (P = 0.002).
|
) was lower in the pigs fed the FLF diet (16 mmol•kg-1•h-1) compared with the pigs fed the DF and NFLF diets (36 to 37 mmol•kg-1•h-1) (P = 0.08). In general, the values decreased in the proximal small intestine and increased along the small intestine. In the cecum and colon contents, no production of lactic acid was detected. The production of acetic acid was also lower in the stomach of pigs fed FLF (2 mmol•kg-1•h-1) compared with those pigs fed the other two diets (14 to 18 mmol•kg-1•h-1) (P = 0.004), whereas similar values were observed in the rest of the GI-tract of all pigs. The production of acetic acid was lowest in the proximal small intestine (2 to 3 mmol•kg-1•h-1), and increased along the GI-tract to values of 10 to 17 mmol•kg-1•h-1 for all diets. The production of propionic acid was lowest in the pigs fed FLF (2 mmol•kg-1•h-1) followed by those fed NFLF (11 mmol•kg-1•h-1) and those fed DF (15 mmol•kg-1•h-1) (P = 0.04). The values varied between 1 and 5 mmol•kg-1•h-1 in the small intestine, and between 4 and 6 mmol•kg-1•h-1 in the cecum and colon for all diets. Butyric acid production was similar for all diets along the GI-tract, with the lowest values measured in the stomach and small intestine (1 to 2 mmol•kg-1•h-1) and the highest in the cecum and colon (3 to 5 mmol•kg-1•h-1) for all diets.
|
). The pigs fed DF grew 961 g/d, which had a tendency (P 0.09) to be different to the growth of the other two dietary groups. The ADFI (on DM basis) was highest for the pigs fed NFLF (2.14 kg/d) followed by those fed DF (2.06 kg/d), and lowest for those fed FLF (1.96 kg/d) (P < 0.001). Gain to feed was similar for all three groups, ranging between 465 and 474 g/kg.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
measured similar starch levels in NFLF and FLF prepared in the same way as in our study. However, Scholten et al. (2001) observed a decrease in the content of starch when grower and finisher liquid diets were fermented during 72 h. The diet composition of Scholten et al. (2001) was different from that used in the present study, which might partially explain the differences observed. Furthermore, in contrast to our procedure, fresh feed was not supplied during the fermentation period, which means that no fermentation substrates (including easily fermentable carbohydrates) were added to the system. This would mean that other substrates (i.e., starch) could then be fermented.
Although only a limited number of samples was analyzed, and the variation of the results was high, lysine concentration seemed to decrease in the FLF compared to DF. Pedersen (2001) reported degradation of 25 to 28% of total dietary lysine during fermentation of FLF. It has been shown that bacteria, such as Salmonella typhimurium (Park et al., 1996) and Escherichia coli (Meng and Bennett, 1992), possess an adaptive response to acid involving a lysine decarboxylase system, which converts lysine to cadaverine. The cell uses these AA decarboxylases to buffer its surroundings by decarboxylating an AA available in the environment, forming a highly basic amine, which is then secreted into the medium. Lysine is added to the feed as free AA, which can make it more susceptible to decarboxylation by the bacteria than other AA bound to proteins. This process could explain the decrease of lysine in the FLF observed in this and other studies (Pedersen, 2001). In disagreement with these findings, preliminary results from our laboratory showed no disappearance of total lysine or free lysine in liquid feed following addition of free lysine and incubation for 168 h at 20°C. So, this phenomenon needs further study in order to elucidate its importance and dependence on various variables related to preparation of FLF, such as indigenous microflora in the ingredients, microflora in the environment, temperature of incubation, and so on.
An important observation regarding the NFLF is that the decrease in LMW sugars and the increase in the levels of lactic acid bacteria, yeasts, and total anaerobes, compared with DF, clearly suggest that fermentation took place in the NFLF. However, the high levels of enterobacteria compared with DF and FLF indicate that fermentation did not reach steady state, and that the feed was in what has previously been defined as the first phase of fermentation. A possible explanation is that the buckets where the feed and water were mixed before being offered to the animals were contaminated with feed from the previous feedings. This residue would act as a "starter culture" and accelerate the fermentation in the feed. Since low pH, high lactic acid bacteria counts, low enterobacteria counts, and high lactic acid concentration are believed to contribute to a more healthy GI-tract of pigs and piglets (Jensen and Mikkelsen 1998; Geary et al., 1999; van Winsen et al., 2001), the high enterobacteria count in the NFLF makes it theoretically less suitable for pig/piglet feeding, at least if health promotion is the aim. Furthermore, the characteristics of the NFLF suggest that under farm conditions (due to contamination of equipment, etc.), even if fresh feed and water are mixed and immediately given to pigs, the lag between feeding and actual intake of the liquid feed could result in feed of undesirable properties with regard to gut health, which agrees with the results of van der Wolf et al. (1999).
Digesta from the Gastrointestinal Tract
The effect of the experimental treatments on enterobacteria is of importance, given that one crucial benefit expected from feeding FLF is the reduction of enterobacteria along the GI-tract of the animals (Mikkelsen and Jensen, 1997; Moran, 2001; van Winsen et al., 2001). The results of the present study showed that FLF (low pH, high density of lactic acid bacteria, high concentration of organic acids, low density of enterobacteria) was efficient in decreasing enterobacteria counts along the entire GI-tract of pigs, and that feeding liquid feed, which was in the initial phase of fermentation (high pH, low density of lactic acid bacteria, low concentration of organic acids, high density of enterobacteria), resulted in higher counts of enterobacteria in the entire GI-tract of pigs than feeding DF. Similarly, Mikkelsen and Jensen (1997) measured higher counts of coliform bacteria in the GI-tract of piglets fed NFLF than in the GI-tract of piglets fed FLF. Data from Moran (2001) did not show higher counts of coliforms in the GI-tract of piglets fed NFLF compared with those fed DF. The levels of coliforms in the small and large intestine of the animals fed DF were much higher (8.4 to 8.7 cfu/g) in Moran’s study than in ours (3.8 to 6.2 cfu/g), which can make increases of the counts more difficult to observe. Our and Mikkelsen and Jensen’s (1997) results do not necessarily mean that healthy and robust animals will suffer gastroenteric disorders if fed NFLF, as supported by the health status and growth performance of the animals in the present study, but it might predispose those already weakened due to other nutritional or management practices to diarrhea, among other things. Studies by van der Wolf et al. (1999) revealed that feeding a complete liquid feed containing fermented byproducts from human food production and participation in an integrated quality-control program were associated with a decreased risk for Salmonella infection, whereas feeding pigs soaked compound feed in a trough increased the risk of infection. Furthermore, van der Wolf et al. (2001) measured a lower Salmonella prevalence in finishing pigs from herds feeding FLF than in those from herds fed pelleted compound feed.
Lower pH and higher concentration of organic acids in stomach contents of pigs fed with FLF compared with animals fed with DF or NFLF, as measured in the present study, have repeatedly been reported in piglets (Mikkelsen and Jensen, 1997; van Winsen et al., 2001; Scholten et al, 2002). A low pH in combination with high concentration of organic acids can impair bacterial metabolism (Russell and Diez-Gonzalez, 1998). Gram-positive bacteria (i.e., lactic acid bacteria) can tolerate this environment better than coliform bacteria, both in the GI-tract (present study; Mikkelsen and Jensen 1997; van Winsen et al., 2001) and in vitro (Geary et al., 1999; van der Wielen et al., 2001; Knarreborg et al., 2002). It is not only the high concentration of organic acids that results in impairment of bacterial metabolism, but also the high concentration in an undissociated form (Russell and Diez-Gonzalez, 1998; van Winsen et al., 2001). The process by which undissociated organic acids, present in the gut, can penetrate into the bacterial cytoplasm is well described. There, they contribute to a decrease of the internal bacterial pH and a collapse of the proton motive force, resulting in impairment of bacterial cell metabolism and/or death (Russell and Diez-Gonzalez, 1998). The concentration of undissociated lactic acid in the stomach of pigs fed FLF was between four- and sevenfold that in the stomach of the other pigs. This indicates that the level of undissociated lactic acid is one of the variables contributing to a decrease of enterobacteria counts in the GI-tract of the pigs fed FLF.
Another mechanism underlying the inhibition of enterobacteria as a result of feeding FLF could be related to a more constant low pH during the day in the stomach of pigs fed FLF compared with a more fluctuating gastric pH in the animals fed the other diets. Data from our laboratory (Canibe and Jensen, 2002) showed that feeding a diet with added 1.8% formic acid maintained gastric pH below 4 for the whole day, which would kill enterobacteria (Knarreborg et al., 2002). However, feeding diets without addition of formic acid resulted in pH values of 4.7 shortly after feeding and in bactericidal levels (pH 4) only after 2 to 4 h postfeeding. In the latter situation, enterobacteria would be able to proliferate in the stomach during the first hours postfeeding and move further down in the GI-tract.
The lower counts of lactic acid bacteria in the cecum and mid-colon and of total anaerobes in the mid-colon of the animals fed FLF vs. those fed DF or NFLF could be explained by a reduction of available substrates for microbial fermentation in these segments of the GI-tract after feeding FLF. The reduced substrate availability for microbial growth in the large intestine may be the result of a virtual absence of LMW sugars in the FLF (as shown in Table 2
) and a higher ileal digestibility of nutrients in the FLF, as suggested by Urlings et al. (1993). However, the similar SCFA production and ATP concentration in the large intestine in all groups does not point to a lower microbial activity in the large intestine of pigs fed FLF. van Winsen et al. (2001) measured higher pH values in the rectum of pigs fed FLF than in those fed DF, which may suggest lower microbial activity in the rectum of pigs offered FLF. However, in agreement with the present study, they found similar concentrations of SCFA in both groups of animals. An explanation for the mentioned inconsistencies has not been found. Højberg et al. (2001) fed the same DF and FLF used in the present study to growing pigs and measured lower fermentation capacity (measured with the Phene-Plate System, Sweden) in the cecum and colon of pigs fed FLF than in those fed DF, which agrees with the lower microbial counts observed in the present study.
There were higher counts of lactic acid bacteria grown at 20°C in NFLF and FLF than in DF. These bacteria were present at a higher level in the proximal GI-tract of pigs fed NFLF or FLF compared with those fed DF, but the levels in the distal GI-tract were similar for all experimental groups. This could indicate that the origin of the lactic acid bacteria grown at 20°C measured in the GI-tract was the feed. This is supported by the lower production of lactic acid in gastric contents of FLF fed-pigs compared with the other pigs (Table 9). Furthermore, the density of these bacteria in the NFLF and FLF groups was higher in the stomach and small intestine (from <6.5 to 9.0 log cfu/g) than in the cecum and colon (from <6.0 to <6.6 log cfu/g), which suggests that lactic acid bacteria grown at 20°C are probably transient but not active in the GI-tract. Differences between dietary groups in counts of lactic acid bacteria grown at 20°C in the proximal GI-tract and in counts of lactic acid bacteria grown at 37°C in the distal GI-tract suggest that the composition of the population of lactic acid bacteria was different in the GI-tract of the pigs fed DF, NFLF, and FLF (Jensen and Mikkelsen, 1998). Counts of yeasts grown at 20°C were highest in the FLF and along the whole GI-tract of pigs fed FLF compared to the other two experimental groups. The role of these microorganisms in the GI-tract of pigs is unknown; therefore, it is difficult to predict whether the stimulation of yeasts by feeding FLF is beneficial or detrimental for the animals. Studies are needed to identify microorganisms present in the NFLF and FLF to distinguish between the bacteria and yeasts of dietary and intestinal origin present in the GI-tract of pigs fed NFLF and FLF, and their metabolic activity in the GI-tract.
The gastric concentration of lactic acid was highest (not significant) in the animals fed FLF. However, the production (mmol•kg-1•h-1) was lowest (not significant), which indicates that the lactic acid measured in the stomach of animals fed FLF was produced in the feed before ingestion, and not in the stomach of the animals. Data from the present study is not conclusive because nonsignificant differences were observed, and a greater number of animals would be needed. However, Jensen and Mikkelsen (1998) drew similar conclusions when comparing gastric contents from piglets fed FLF and NFLF, which supports our hypothesis. van Winsen et al. (2001) observed a higher gastric concentration of lactic acid in piglets fed FLF compared to those fed DF, but production was not measured. van Winsen et al. (2001) suggested that the most important variables of feeding FLF that contribute to the reduction of undesirable microorganisms in the GI-tract are pH and organic acid. In this study and in others (van Winsen et al., 2001; Scholten et al., 2002), feeding FLF influenced pH and organic acid concentration mainly in the stomach. However, differences in microorganism counts are often detected all along the GI-tract. This indicates that changes in gastric contents that reduce pathogen survival or proliferation in the stomach also seem to reduce the presence or proliferation of pathogens along the rest of the GI-tract. This suggests that the stomach acts as a barrier against colonization of pathogens in the GI-tract and that this site should be considered to be a crucial site in the control of undesirable microorganisms/pathogens in the GI-tract of pigs. If this hypothesis is correct, feeding/management strategies aimed at reducing pathogen survival in the GI-tract of pigs should focus on reducing survival of such microorganisms in the stomach.
Growth Performance
Jensen and Mikkelsen (1998) concluded in their review that feeding liquid feed to slaughter pigs seems to improve the efficiency of feed utilization, whereas the effect on growth rate, compared to DF, is less consistent. In the present study, gain:feed ratio was similar for liquid and dry feed. Because fermentation can start from the moment water and feed are mixed, the various procedures used to make NFLF may result in different degrees of fermentation of the feed, which leads to feed of different characteristics.
Data on the effect of feeding NFLF or FLF to growing pigs on growth performance are scarce (reviewed by Jensen and Mikkelsen, 1998; Hurst et al., 2001). Most studies have investigated the effect on piglets (Blanchard et al., 2000; Moran, 2001; Scholten, 2001). In accordance with the present results with growing pigs, Moran (2001) measured higher feed intake and weight gain, and similar feed to gain ratio in piglets fed NFLF compared to feeding FLF. Pedersen (2001) measured higher growth performance in piglets fed NFLF vs. those fed DF or FLF, and similar values for those fed DF and FLF. Disappearance of lysine in the FLF due to fermentation was suggested by Pedersen (2001) as the main reason for the negative effect of FLF compared to NFLF. In our study, the similar gain:feed ratio observed for these diets do not point to lysine as reaching values below needs of the FLF-fed animals. A study with growing pigs carried out by Hurst et al. (2001) showed better FCR when feeding liquid feed compared to feeding DF. The effect on daily feed intake and weight gain depended on the feed:water ratio of the liquid feed, but in general, a lower feed intake and higher weight gain was the reason for the improved feed:gain ratio. Lawlor et al. (2002) measured greater DMI in pigs fed FLF compared with pigs fed dry-pelleted feed from 0 to 28 d postweaning, whereas gain:feed ratio was lower for the piglets offered FLF. A too-low pH (Danish field studies suggest that pH values of 4.2 could be too low in some cases) due to overfermentation in the FLF could result in depressed feed intake by the animals. However, the pH values of the FLF used in the present study (4.4 ± 0.2) do not point to overfermentation as the reason for the decreased feed intake. On the other hand, end metabolites and acidity of the liquid feed following fermentation may impair palatability (Moran, 2001).
|
Implications |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
1 Correspondence: P.O. Box 50 (phone: +45-89-99-11-48; fax: +45-89-99-13-78; E-mail: nuria.canibe{at}agrsci.dk).
Received for publication July 19, 2002. Accepted for publication April 25, 2003.
|
Literature Cited |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. Olstorpe, K. Lyberg, J. E. Lindberg, J. Schnurer, and V. Passoth Population Diversity of Yeasts and Lactic Acid Bacteria in Pig Feed Fermented with Whey, Wet Wheat Distillers' Grains, or Water at Different Temperatures Appl. Envir. Microbiol., March 15, 2008; 74(6): 1696 - 1703. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Canibe, O. Hojberg, J. H. Badsberg, and B. B. Jensen Effect of feeding fermented liquid feed and fermented grain on gastrointestinal ecology and growth performance in piglets J Anim Sci, November 1, 2007; 85(11): 2959 - 2971. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Ogino, H. Hirooka, A. Ikeguchi, Y. Tanaka, M. Waki, H. Yokoyama, and T. Kawashima Environmental Impact Evaluation of Feeds Prepared from Food Residues Using Life Cycle Assessment J. Environ. Qual., May 25, 2007; 36(4): 1061 - 1068. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Canibe, O. Hojberg, S. Hojsgaard, and B. B. Jensen Feed physical form and formic acid addition to the feed affect the gastrointestinal ecology and growth performance of growing pigs J Anim Sci, June 1, 2005; 83(6): 1287 - 1302. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. G. Manzanilla, J. F. Perez, M. Martin, C. Kamel, F. Baucells, and J. Gasa Effect of plant extracts and formic acid on the intestinal equilibrium of early-weaned pigs J Anim Sci, November 1, 2004; 82(11): 3210 - 3218. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS |
) and fermented liquid feed (
). aSto = stomach; SI1 = proximal small intestine; SI2 = mid-small intestine; SI3 = distal small intestine; Ce = cecum; Co1 = proximal colon; Co2 = mid colon; Co3 = distal colon. Values are least square means and their standard error (n = 5). a,bSignificant difference (P < 0.05).