Materials and Methods

Experiment 1—Four spayed-female mixed-breed dogs were used to determine the optimal window for flatulence monitoring and to determine if the number of feedings per day alters flatulence. Initially, all dogs were fed twice daily at 7:00 AM and 5:00 PM for 1 week. In vivo flatulence was monitored⁵ for three 24-hour periods while dogs were maintained on a commercially available adult dog food. During the second half of the experiment, all dogs were fed once daily at 7:00 AM for 1 week. After 1 week of feeding, in vivo flatulence was monitored for three 24-hour periods.

Experiment 2—Six adult spayed-female mixed-breed dogs weighing 18.8 ± 0.2 kg (mean ± SEM) were used in a 6 × 6 Latin square design experiment with a 2 × 3 factorial treatment structure to evaluate the digestibility, flatulence, and fecal odor metabolites of LLM, SBM, and PBP meal–based foods with or without supplemental β-mannanase^d (5 g/kg). Dogs were housed in the Division of Laboratory Animal Research Facility at the University of Kentucky and were cared for in accordance with Institutional Animal Care and Use Committee–approved protocols. The dog's primary living space was cleaned twice daily after feeding times. Dogs were exercised daily and environmental enrichment was provided, which included, but was not limited to, cage play (toys), grooming, and other human-dog interactions (ie, petting). Water was available ad libitum throughout the entire experiment.

Feeding and treatments—The ingredient and nutrient compositions of each food treatment are depicted (Tables 1 and 2). Each food was kibbled and formulated in accordance with the Association of American Feed Control Officials⁹ nutrient guide for dogs and balanced to meet adult maintenance requirements. The enzyme, β-mannanase, was provided by adding 5 g of β-mannanase/kg of diet prior to feeding. This concentration of supplemental β-mannanase was determined as a result of the highest reported concentration of β-mannan detected in SBM (15 g/kg of SBM).^7,c Each day, food and enzyme was weighed and fed at 7:00 AM in stainless steel bowls. Each dog was allowed 20 minutes to consume the food. Bowls were removed after 20 minutes, and orts were weighed and recorded. Throughout the experiment, food samples were collected daily and pooled into plastic collection bags for nutrient content analysis.

Sample collection—The duration of each experimental period was 14 days. During the first 2 days of the period, dogs were transitioned to the next diet by feeding a 1:1 mixture of their current diet and the new diet. Dogs were allowed 6 days for adaptation to the test diet. All dogs were fed once daily at 7:00 AM.

Fecal collection periods consisted of 5 days (days 6 to 11). Fecal collection began on day 6 at 5:00 PM. All feces present prior to this first collection were discarded. All feces from 5:00 PM on day 6 through the end of the 5-day period were collected into labeled plastic bags. Fecal scores were determined at each collection on a scale from 1 to 5 for volume, stickiness, adhesiveness, and moisture. All feces collected during the first 2 days of fecal collection were stored at 4°C for fecal odor metabolite analysis. All feces collected during the last 3 days of fecal collection (days 9 to 11) were weighed and pooled by dog within each fecal collection period. On the days of in vivo flatulence measurement (days 12 to 14), all dogs were fed at 7:00 AM. When each dog was finished eating, a vest was placed on the dog. Flatulence was monitored with a gas monitor^e beginning 2 hours after feeding.⁵ This pump monitored the emission of H₂S rectal gases for 12 hours. These data were used to determine episodes of flatulence per day (number of peaks), the total amount of H₂S produced during monitoring, and the peak concentration of H₂S produced during the monitoring period (max H₂S).

Analyses—Fecal samples stored at 4°C were analyzed for fecal odor metabolites by use of SPME. Fecal samples were collected and stored at -20°C prior to further handling. Samples were thawed for 12 hours at 22°C and homogenized prior to measuring into vials. Five grams of each sample was expressed into a 20-mL headspace vial through a 10-mL polyethylene syringe. Vials were capped by use of polytefcoated septa and crimp caps to maintain an airtight seal. Vials were stored at 4°C until they were entered into the gas chromatograph mass spectrophotometer queue. Each sample was prepared in duplicate, and the mean results were calculated.

The SPME was performed by use of a preconditioned 2-phase fiber.^f The SPME procedure was performed by use of a robotic system.^g Each sample was equilibrated at 30°C for 5 minutes prior to SPME exposure. The fiber was exposed to the headspace of each vial in turn for 30 minutes at 30°C. After exposure, the fiber was inserted into the injection port of the gas chromatograph mass spectrophotometer for 5 minutes.

A mass selective detector^h interfaced with a gas chromatographⁱ was used to perform the separation and subsequent mass analysis. The separation was performed by use of a capillary column.^j The capillary injector was held at 310°C. Helium was used as the carrier gas. Injections were made with a 50:1 ratio split with constant linear flow velocity of 39 cm/s. The initial oven temperature of 60°C was held for 2 minutes after the injection. The temperature was ramped up at 5°C/min to 150°C, held for 2 minutes, and then ramped up at 8°C/min to 250°C. This final temperature was held for another 23.5 minutes. The mass selective detector scanned a range from 35 to 220 atomic mass units. Compounds in the effluent were identified by comparison to retention times and fragmentation patterns of known materials or by comparison to fragmentation patterns in a mass spectral library.^k

After collection, the last 3 days of fecal samples were stored at -20°C until they were lyophilized.^l Dry matter was determined by weighing each sample before and after lyophilization. Fecal samples were ground through a 0.5-mm screen in a sample mill.^m Feed samples were ground by use of a blender.ⁿ The dried and ground samples were stored in labeled plastic bags at 22°C until further analysis.

Protein content of feed and fecal samples was obtained (nitrogen × 6.25) by use of a nitrogen analyzer.^o Energy content of feed and fecal samples was obtained by use of a bomb calorimeter.^p

Sucrose, raffinose, and stachyose were determined in food samples by use of the following procedure. Ground samples (0.5 to 1.0 g) were placed into a screw cap test tube to which 10 mL of 50/50 (vol/vol) ethanol-to-water was added. Samples were mixed well, and test tubes were placed into a shaker bath at 40°C for 30 minutes. Samples were then placed into a sonicator for 10 minutes. Samples were placed into a shaker bath for an additional 30 minutes. After shaking, samples were centrifuged at 2,000 × g for 15 minutes and 1 mL of supernatant was transferred into a microcentrifuge tube. Samples were dried by use of a rotary evaporator (3 to 4 hours) and derivatized, as described by Knudsen and Li.¹⁰ The resulting solution was analyzed by use of a gas chromatograph^q with an autosampler.¹¹

Statistical analysis—Data were analyzed as a 2 × 3 factorial treatment structure in a 6 × 6 Latin square by use of ANOVA and general linear models.^r The experimental unit was dog, and the model included period, dog, protein source, and supplemental enzyme. Preplanned comparisons were used to compare treatment means for protein sources (PBP vs SBM plus LLM, and SBM vs LLM). A value of P < 0.05 was considered significant.

Results

All dogs remained healthy throughout both experiments. Results of experiment 1 are depicted (Table 3). The data suggested that when feeding once daily, the optimal time frame to monitor flatulence was approximately 2 hours after feeding. In experiment 2, no differences in body weight were observed (mean, 18.8 kg; Table 4). This was expected because intake was adjusted to supply the proper amount of energy required for maintenance of body weight at the beginning of each experiment. Because intakes were adjusted each period to maintain body weight, there were no differences in DM intake (mean, 309 g of DM/d), nitrogen intake (mean, 9.8 g of nitrogen/d), or energy intake (mean, 4,721.7 cal/g). Supplemental β-mannanase had no effect on total tract DM and nitrogen digestibilities or digestible energy; however, differences between protein sources did exist for total tract DM digestibility and digestible energy. Poultry by-product had higher DM digestibility and digestible energy (0.913 and 4,255 cal/g), compared with both soy products (mean, 0.870 and 4,049 cal/g), and LLM had a higher DM digestibility and digestible energy than SBM.

Fecal output on an as-is basis was lower (mean, 66.3 g/d; P < 0.001) for dogs consuming PBP, compared with the soy-fed dogs (mean, 134.7 g/d; Table 5). No differences were detected in total fecal output of SBM- or LLM-fed dogs. Fecal moisture was lower (mean, 599 g/kg; P < 0.001) in PBP-fed dogs, compared with the mean of soy-fed dogs (698 g/kg). No differences in fecal moisture were detected when SBM was compared with LLM. Fecal quality was determined by visual scoring for volume, stickiness, adhesiveness, and moisture. Dogs consuming PBP foods had lower volume (mean, 2.6; P = 0.001), stickiness (mean, 2.3; P < 0.001), and adhesiveness (mean, 3.6; P = 0.002), compared with the mean of all soy foods. No differences were detected in fecal scores when SBM and LLM foods were compared.

No differences were detected for all fecal metabolites regardless of the addition of supplemental enzyme; however, differences among protein sources were detected (Table 6). The PBP diet had lower concentrations of fecal carboxylic acids and esters, compared with SBM and LLM. Dogs had higher concentrations of heterocycles (P < 0.001), phenols (P < 0.001), thio and sulfides (P<0.001), ketones (P<0.001), alcohols (P<0.001), and indoles (P<0.001) when fed PBP than when fed the soy-containing foods. No differences were detected in any fecal odor metabolites with the exception of carboxylic acids when SBM was compared with LLM. The SBM had the highest concentration of carboxylic acids, compared with all other treatments. No differences were observed among treatments for any flatulence components analyzed (Table 7).

Discussion

The primary objective of the study reported here was to evaluate the effects of oligosaccharides and β-mannans in soybean meal on nutrient availability and flatulence in dogs. Soybean meal is often integrated into animal foods because it is an excellent source of protein and digestible energy; however, the inclusion of soybean meal into diets for dogs and other monogastric animals has been linked with an increase in flatulence and a reduction in digestibility because of the presence of oligosaccharides.^1,2 Oligosaccharides of the raffinose family, predominantly found in legumes, are believed to be the major factor influencing flatulence and are 1 factor responsible for the reduction of digestibility in monogastrics.

The 2 predominant oligosaccharides found in soy are stachyose and raffinose. Stachyose and raffinose are believed to lower digestibility because they are not digested in the small intestine. These sugars cannot be cleaved in the small intestine because of the absence of α-1,6-galactosidase in the small intestinal tract.¹² As a result, virtually all of these sugars pass intact into the large intestine, where they are hydrolyzed and fermented by bacteria into short-chain fatty acids and gases causing flatulence.

Stachyose and raffinose concentrations in soybean meal reportedly range from 32 to 52 g/kg and 6 to 14 g/kg of DM, respectively.^3,a However, results of a previous study² indicate that concentrations of stachyose may be as high as 112 g/kg of DM in soybean meal. This high concentration of stachyose in soybean meal may be the cause of decreased small intestinal digestibility.² Zuo et al¹² investigated the effects of oligosaccharide on soybean digestibility in dogs. In that study, dogs were fed diets containing 180 and 370 g of conventional soybean meal or low-oligosaccharide soybean meal per kg. The resulting diets contained concentrations of stachyose ranging from 0 to 27 g/kg of DM. Results of that study indicate that oligosaccharides were not responsible for the reduction in nutrient digestibility when dogs are fed diets containing <27 g of stachyose/kg. Yamka et al² suggested that a diet containing <30 g of stachyose/kg does not result in a reduction of nutrient digestibility. When dogs were fed diets exceeding these concentrations (30 to 50 g/kg of DM), Yamka et al² reported lower digestibility of SBM in dog diets. In the study reported here, the presence of oligosaccharides in SBM decreased total tract DM digestibility (approx 2 percentage units) and digestible energy (approx 100 cal/g), compared with LLM.

Suarez et al⁵ studied gas production and gastrointestinal tract symptoms in humans consuming conventional soy flour (33.3 g of stachyose/kg and 5.1 g of raffinose/kg) and low-oligosaccharide soy flour (4.6 g of stachyose/kg and 1.6 g of raffinose/kg). They found that humans consuming conventional soy had significantly higher breath hydrogen concentrations (20 μg/mL), compared with humans consuming low-oligosaccharide soy and white-rice cereal (control). The increase in breath hydrogen was believed to be the result of increased fermentation in the large intestine because of the presence of oligosaccharides. Consumption of conventional soy also resulted in a higher flatus frequency (7.5 vs 3.8 times/12 h), compared with the low-oligosaccharide soy and the control. Although frequency of flatus was reported, no measures of compounds causing flatus odor (H₂S) or its intensity were determined.

A method for in vivo assessment of flatulence in dogs has been described.⁶ This method was applied in our study to determine whether dietary factors affect flatulence. Although this method appeared to be effective in monitoring flatulence, the optimal window for collection was never determined. Collins et al⁵ based their collections on results of studies in humans (optimum 4 hours after feeding). In experiment 1, we found that dogs first had flatulence approximately 2 hours after feeding. We also found that feeding dogs once daily, compared with twice daily, had an effect on flatulence, and typically, dogs had 3.5 more episodes/d when fed once daily. In the study reported here, oligosaccharide content of diets did not alter any of the variables investigated, although there was a trend for dogs receiving supplemental β-mannanase to have a higher number of episodes of flatulence, higher total amount of H₂S produced, and higher peak concentrations of H₂S than dogs that did not receive supplemental β-mannanase. The addition of β-mannanase could have resulted in increased substrate being available for microbial populations in the large intestine. Unfortunately, we were not able to detect differences between treatments given the high day-to-day variability in flatulence.

Although results of studies^2,3,12 investigating the role of oligosaccharides in digestibility are available, studies concerning other components of the carbohydrate portion of soybean meal are limited. Soybean meal can contain 12.5 to 15.0 g of β-mannans/kg, which are indigestible and heat stable.^b β-Mannan is a linear polysaccharide composed of repeating β-1-4 mannose and α-1-6 galactose and glucose units attached to a β-mannan backbone. The presence of β-mannans in soy-based diets results in lower feed conversion for swine⁷ and laying hens.⁸ Jackson et al⁸ found that administering β-mannanase to laying hens fed cornsoy diets resulted in increased egg weight. The addition of 0.5 g of β-mannanase/kg to swine diets increases growth performance in weanling and growing-finishing pigs; however, no differences were detected in energy, nitrogen, or DM digestibility.⁷ Radcliffe et al^c gave swine supplemental β-mannanase (5 g/kg), and an increase in small intestinal DM digestibility and energy availability was detected. In our study, supplemental β-mannanase (5 g/kg) had no effect on DM, nitrogen, or energy digestibilities. In our study, if β-mannans were present in diets at the highest reported concentrations (15 g/kg of DM), the supplemental enzyme would not likely have a measurable effect on digestibility because SBM and LLM were less than a third of the diet (β-mannans maximally contribute 5 g/kg of diet). However, the effects of β-mannans on flatulence are not clear. This suggests that β-mannans are not likely to alter the digestive capabilities of the adult dogs fed at maintenance. These results are in agreement with results of studies^7,c indicating that supplemental β-mannanase has minimal effects on nutri-ent digestibility and growth performance in weanling and growing swine.

Odor components in feces from dogs and other monogastrics are the end products of fermentation. They are produced by bacterial enzymes during the degradation of endogenous and undigested food proteins.¹³ The fecal odor components include ammonia, aliphatic amines, indoles, phenols, and volatile sulfur-containing compounds. Odor components not only affect fecal odors but can also cause serious negative effects on the gastrointestinal health of animals.¹³ These fecal components can be altered by the amount of dietary amino acids that reach the large intestine. This can be affected by high DM intake, high-protein diets, and diets that contain low-quality proteins. An increase in protein and amino acid flow to the large intestine results in more substrate for resident bacteria and an increase in fecal odor components.¹³ Many studies investigating fecal odor components have focused on the use of prebiotics in altering the fecal odor compounds. Hussein and Sunvold¹³ found that feeding adult dogs fructo-oligosaccharides (0.5% of the DM) resulted in a decrease in putrescine, cadaverine, tyramine, and total amines. Results of that study indicate that fecal odor can be manipulated by prebiotics; however, few studies have focused on the association between fecal components and protein source.

In the study reported here, all protein sources had similar total tract nitrogen digestibility. Nevertheless, the animal protein source caused high concentrations of heterocycles, phenols, thio and sulfides, ketones, alcohols, and indoles. Although dogs fed PBP meal had low fecal output, the fecal odor metabolites excreted per day were greater for dogs consuming PBP than those consuming LLM and SBM. These data suggest that dogs fed PBP had feces that contained more unpleasant odor components than soy-fed dogs. The reason for this difference is not clear because the amount of nitrogen actually entering the large intestine for fermentation in this study was not known. Results of a previous study⁶ indicate that dogs fed low-ash poultry meal and SBM had similar small intestinal nitrogen digestibility; however, large intestinal nitrogen digestibility was approximately 20 percentage units higher in dogs fed low-ash poultry meal. This increase in large intestinal fermentation of nitrogen may be the cause of the high fecal odor compounds detected in our study. The difference in fecal odor metabolites may also have been the result of lower fiber amounts in the PBP diet, compared with the soy-based diets, resulting in greater protein fermentation. The 2 soy-based diets in our study had 2 times the amount of crude fiber as the PBP diet. Additionally, the soluble fibers present in soy diets can act as a prebiotic, which can decrease pH, increase lactobacilli and bifidobacteria, and reduce the numbers of detrimental bacteria, which are associated with fecal odor compounds.

Results of our study indicated that LLM and SBM can be good sources of protein for dogs and have a high digestibility similar to PBP. No significant differences were detected among treatments for flatulence because of the high day-to-day variability. The presence of oligosaccharides in SBM decreased total tract DM digestibility (approx 2 percentage units) and digestible energy (approx 100 cal/g), compared with LLM. Dogs fed PBP had high concentrations of fecal odor compounds, indicating that a high amount of nitrogen fermentation was occurring in the large intestine.

Results of the study reported here suggested that foods containing <22.4 g of stachyose/kg and <2 g of raffinose/kg did not alter digestibility or increase flatulence in dogs. Addition of supplemental β-mannanase to soy-based diets did not enhance or alter nutrient digestibility, indicating that β-mannans may not be detrimental in these diets. Results also suggested that oligosaccharides and β-mannans are not responsible for altering nutrient digestibility and increases in flatulence. Other factors or soluble fiber present in soy products may be the potential initiator for decreased digestibility and flatulence. Foods containing soybean meal may be beneficial in reducing fecal odor compounds in dogs with fecal odor problems.