| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
U.S. Meat Animal Research Center, ARS, USDA, Clay Center, NE 68933-0166
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResults and DiscussionImplicationsLiterature Cited |
|---|
Key Words: Bacteria • Manures • Odors • Swine
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionImplicationsLiterature Cited |
|---|
; Hartung and Phillips, 1994) produced during digestion and subsequent manure storage. Microorganisms mediate odor compound production during the incomplete anaerobic fermentation of substrates in manures (Mackie et al., 1998). Aromatic compounds and VFA most closely correlate to swine odor (Zahn et al., 1997; 2001; Powers et al., 1999); therefore, limiting odor precursors (microbial substrates) in manures would limit odor emissions.
Microorganisms may use a variety of potential substrates, including starch, proteins, lipids, and nonstarch polysaccharides (NSP), to produce odor compounds (Mackie et al., 1998; Zhu et al., 1999). In cattle manures, starch seems to be preferred even when other substrates are available (Miller and Varel, 2001; 2002); addition of starch to cattle manures prevented protein fermentation, and no malodorous branched-chain VFA or aromatic compounds accumulated. Odor compound accumulation in cattle manure appears to be self-limiting due to lactate accumulation and pH-associated VFA toxicity. Although there is good evidence for a link between swine manure odor and diet (Sutton et al., 1999), the biochemical sources of odors are not well characterized. Clarifying the role of starch, protein, and cellulose in the production of odor compounds from swine manure should aid the formulation of better, low-odor swine diets.
The objective of this study was to identify the contribution of different substrates in swine manure slurries that are fermented to odorous compounds. Both the disappearance of potential substrates and the production of fermentation products were monitored in swine manure slurry incubations over 5 wk. Results were compared to an earlier experiment of cattle feedlot manure and illustrated similarities and differences in the biochemical sources and microbial pathways for odor compound production in animal manures.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionImplicationsLiterature Cited |
|---|
For manure slurries, 0.8 kg of the fresh swine manure was blended in a Waring blender (New Hartford, CT) with 3.2 L of H2O to form a 20% (wt/vol) manure slurry. The initial percentage weight composition of the swine manure composite was 27% DM, 22.6% OM, 0.9% nonammonia nitrogen (NAN), 1.5% starch, and 11.6% NSP (methods of analyses described below). The slurry was equally divided into four blenders with four different substrate treatments (protein, starch, cellulose, and no addition). The substrate concentrations added to three different blenders were casein (Fisher Scientific, Fair Lawn, NJ) at 5 g/L, starch (Sigma, St. Louis, MO) at 2 g/L, and microcrystalline cellulose (J. T. Baker, Phillipsburg, NJ) at 6 g/L. The quantities of the additions were designed to increase the endogenous concentration of a particular substrate by 50%. Each sample was blended and approximately 190 mL of manure slurry was then added to five 250-mL flasks (five replicate flasks per manure/treatment). Flasks were gassed with N2, stoppered to limit volatilization losses and ensure anaerobic conditions, and incubated at room temperature (20 to 23°C). Excess fermentation gas was vented though a needle into a water-filled test tube. Every day during the first week, slurries were mixed, and 8 mL was collected and distributed into three aliquots (4 mL for pH and substrates, 2 mL for fermentation products, and 2 mL for DM and OM determination). After the first week, samples were collected less frequently (twice to once per week). Manure slurry pH was analyzed immediately using a combination pH electrode, whereas the remaining manure parameters (substrates and fermentation products) were determined from frozen samples analyzed after the end of incubation. Details of the analyses have been previously described (Miller and Varel, 2001; 2002). Briefly, manure substrates, which included NAN, starch, and NSP, were analyzed in homogenized samples. Samples for NAN content, analogous to fermentable nitrogen (protein and nucleic acids), were made alkaline by the addition of 1 mL of 2 M NaOH and dried overnight at 100°C to remove free NH3 before analysis using a Leco CN-2000 carbon/nitrogen analyzer (Leco, St. Joseph, MI). Drying under alkaline conditions removed 100% of NH3 spiked into control manure samples (data not shown). Starch was measured as free glucose using the membrane-immobilized glucose oxidase enzyme system (YSI model 2700, Yellow Springs Instrument Co., Yellow Springs, OH) after overnight digestion with amyloglucosidase. Total polysaccharide was determined colorimetrically using the phenol-sulfuric acid reaction (Daniels et al., 1994), with NSP calculated as the difference between total polysaccharide and starch. Fermentation products (L-lactate, ethanol, propanol, isobutanol, butanol, pentanol, hexanol, acetate, propionate, isobutyrate, butyrate, isovalerate, valerate, isocaproate, caproate, heptanoate, caprylate, phenol, 
-cresol, 4-ethyl phenol, indole, skatole, benzoate, phenylacetate, and phenylpropionate) were quantified in the liquid phase of the slurries using the YSI 2700 analyzer for L-lactate by immobilized L-lactate oxidase enzyme method, and a Hewlett Packard 6890 gas chromatograph (Agilent Technologies, Palo Alto, CA) equipped with flame ionization and mass selective detectors for all other products. Operating conditions used to separate and quantify the fermentation products by gas chromatography have been previously described (Miller and Varel, 2001; 2002).
Odor Source and Substrate Conversion Calculations
Two methods were used to describe the relative contribution of starch or NAN crude protein toward odor compound production in the slurries. The first method calculated the molar percentage of total VFA that was branched-chain VFA in our treatments and compared that percentage to the values reported by Smith and Macfarlane (1996; 1998), who found that purified human colonic bacteria grown anaerobically on protein as the sole carbon source produced up to 17% of the total VFA as branched-chain VFA. The second method calculated the mass loss of starch and NAN crude protein and compared that mass loss to the mass gain of fermentation products in the flask during the incubation period. Percentage conversion of substrates to products was simply:
 |
Statistical Analyses
Data were analyzed as a split-plot in time. The unit of observation was the flask (n = 20; five replicate flasks per treatment) and there were no missing values. The model included effects of incubation day, treatment, flask (treatment), and incubation day x treatment. Treatment was tested against the mean square of flask (treatment). Differences between least squares means were tested with a protected t-test. Statistical analyses were conducted with the GLM procedure of SAS (v. 7.0, SAS Inst., Inc., Cary, NC).
|
Results and Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionImplicationsLiterature Cited |
|---|
). Volatile fatty acids accumulated in all treatments during the first 3 wk (P < 0.001), but the total VFA accumulated varied between treatments (Figure 1A). The starch treatment had greater (P < 0.001) maximum VFA content than all other treatments. Protein and cellulose treatments did not affect (P 
0.156) VFA accumulation relative to the no-addition treatment.
|
). The accumulation of both groups of compounds varied between treatments but produced a similar pattern (Figure 1B, C
). Branched-chain VFA and aromatic compounds accumulated in all treatments (P < 0.001), but the greatest accumulation of these compounds occurred in the protein treatment compared with all other treatments (P < 0.001). The starch amendment accumulated more branched-chain VFA than the no-addition treatment (P = 0.032), whereas no difference was observed between the cellulose treatment and the no-addition treatment (P = 0.592).
Although the concentration of many odorous compounds increased during the incubation, the molar composition (moles/100 moles) of the dominant fermentation products changed very little during the incubation (Table 1). The VFA pool was largely dominated by short, straight-chain VFA (acetate, propionate, and butyrate), which initially comprised 91.5% of the total VFA pool. The percentage of short, straight-chain VFA remained the greatest contributor in the total VFA pool throughout the incubation, but by the end, had decreased in all treatments (P < 0.005). By the end of incubation, the starch treatment had a greater proportion of short, straight-chain VFA than the protein treatment (P < 0.005). Of the three compounds in the short, straight-chain pool, acetate initially comprised a higher molar percentage of the total VFA pool than propionate and butyrate. By the end of incubation, acetate had decreased (P < 0.005) in the no-addition and protein treatments, but it was still the dominant VFA in all the fecal slurries. The percentage of butyrate and propionate either decreased or remained unchanged. The increase in percentage of valerate and branched-chain VFA in all treatments (except branched-chain VFA in the starch treatment) likely accounted for the decreases in short, straight-chain VFA. Substrate addition compared with the no-addition treatment had little effect on the final VFA composition, with the exception of the protein amendment, where the percentage of branched-chain VFA increased relative to straight short-chain VFA. Thus, we concluded that substrate availability in the swine manure affects both the molar concentration of odorous chemical compounds and, depending on the substrate, the final composition of odorous compounds. Furthermore, the range of possible fermentation products will have differing volatilities, odor thresholds, and odor offensiveness, which all impact perceived odor.
|
) and enhance the volatility of acidic fermentation products. In a previous study with cattle manure (Miller and Varel, 2001), lactate was a major fermentation product (concentrations reached 50 mM), which led to very low pH (<4.5) and VFA toxicity to the fermentative microorganisms (Switzenbaum et al., 1990). Lactate accumulated (P < 0.001) in the various swine manure treatments (Figure 2A), but the concentration of lactate was relatively low compared to total VFA accumulation. All substrate additions accumulated more lactate (P 
0.002) than the no-addition treatment. The protein addition, however, had greater lactate accumulation than any other treatment (P < 0.001). The pH of the swine slurries was relatively unaffected by lactate concentration (Figure 2B). The decrease in the pH swine manure slurry was very small (0.3 to 0.6), and the pH never approached the pH 4.75 threshold for VFA toxicity (Switzenbaum et al., 1990; Miller and Varel, 2001). The relative stability of swine manure slurry pH can be attributed to low lactate concentrations in all treatments and to the buffering capacity of ammonia nitrogen normally found in swine manure.
|
), provided indirect evidence for protein fermentation in all treatments and in the no-addition treatment. An estimate of the proportion of protein fermentation relative to starch was made by examining the proportion of branched-chain VFA to total VFA produced exclusively during the incubation. Of the total VFA accumulated during the incubation in the no-addition treatment, 11.7% was branched-chain VFA. The starch and cellulose amendments also accumulated similar percentages (9.4 and 10.5%, respectively) of branched-chain VFA, but did not differ from the no-addition treatment (P 0.21). The protein treatment accumulated more (P 0.025) branched-chain VFA as a percentage of total VFA than all other treatments, averaging 17.3%. Mixed culture microbial fermentations of human colonic bacteria utilizing only protein as a substrate typically produce 7 to 17% of total VFA as branched-chain VFA, with large increases in aromatic compound concentrations (Smith and Macfarlane, 1996; 1998). Based on the percentage of branched-chain VFA in the total VFA pool and the large accumulation of aromatic compounds in our swine manure incubations, we estimate that protein fermentation pathways produced at least half, and maybe substantially more, of the VFA that accumulated during the swine manure incubation. In the protein treatment, protein fermentation may supply essentially all the VFA accumulated during the incubation.
Starch, NAN, and NSP substrate concentrations during the incubation provide a more direct approach to determining the relative contribution of the various substrates to odor compound accumulation during swine manure fermentation. The accumulation of fermentation products, including VFA, alcohols, and lactate, was related to the disappearance of starch and NAN (Figure 3). Correlations were strongest between the accumulation of fermentation products and starch disappearance (r = 0.940, P < 0.001) and NAN disappearance (r = 0.761, P < 0.001) for all data. The correlation between the accumulation of fermentation products and NSP was much weaker (r = 0.280, P = 0.066). All treatments produced similar amounts (P 0.253) of total fermentation product, ranging from 2.8 to 3.4 g/L (± 0.3, n = 20) for the no-addition and starch treatments, respectively. Starch and NAN in all treatments decreased (P < 0.001) during the incubation (Figure 3B, C). Decreases in starch for all treatments ranged from 2.5 to 3.2 g/L (± 0.2, n = 20); the no-addition treatment had the smallest decrease but did not differ from the protein or cellulose treatments (P > 0.162), whereas the starch treatment had a greater decrease compared to the no-addition control (P = 0.006). Decreases in NAN ranged from 0.59 to 1.31 g/L (± 0.08, n = 20); the cellulose treatment had the smallest decrease but did not differ from the no-addition or starch treatments (P > 0.548), whereas the protein treatment had the largest decrease and differed from all the other treatments (P < 0.001). When NAN loss was expressed as CP (NAN x 6.25) loss, CP losses ranged from 3.7 to 8.2 g/L, which equaled or exceeded starch losses. When either starch or protein was added to the slurries, the amount of that particular substrate consumed was greater (P < 0.01) than the consumption in the no-addition treatment. In other words, the greatest losses of a particular substrate were associated with treatments where the substrate was supplied in a highly available form, such as soluble starch or casein. The cellulose treatment was an exception to this trend as NSP consumption only tended to differ (P = 0.075) from NSP consumption in the no-addition control treatment. Although NSP was consumed in the starch, protein, and cellulose treatments (P < 0.03), no NSP was consumed in the no-addition control treatment, thus the contribution of NSP to odor compound formation is more difficult to assess.
|
0.181). Results from substrate losses and percentage of branched-chain VFA in the total VFA pool agree remarkably well. Protein fermentation accounted for half or more of the odor-producing fermentation processes in fresh swine manure and became more important as protein content in the manure increased. The contribution of starch fermentation toward odor compound production, however, should not be overlooked.
Odor Formation Processes Differ Between Cattle Feedlot and Swine Manure
A generally held belief is that odors from swine production are more objectionable than odors from cattle production. Manure storage, handling, and treatment practices are very different between (and within) swine production and cattle feedlot systems and are likely to account for some of the differences in perceived odor. However, odor compounds originating from both systems share a common microbial origin (i.e., microbial fermentation of undigested substrates in manures). A portion of the perceived odor difference between feedlot cattle and swine may be attributable to differences in odor-producing fermentation pathways, which are influenced by differences in substrate availability and/or content in swine and cattle feedlot manures.
The initial substrate composition of fresh swine manure in the control (no-addition) treatment was compared to the substrate composition of fresh cattle feedlot manure in an earlier study (Table 2). The NSP comprised the largest fraction of potential substrate in the swine manure when compared to other substrates on a DM basis, whereas starch was the major substrate in fresh cattle feedlot manure. Starch content in the cattle feedlot manure was nearly sixfold greater than the starch content of swine manure. For NAN (our marker of CP), the situation was reversed, with NAN content in the swine manure twice as high as NAN in the cattle feedlot manure. Our earlier work with cattle feedlot manure indicated that starch fermentation in fresh cattle manures was the dominant route for odor compound formation (Miller and Varel, 2001; 2002). Furthermore, adding starch to aged, starch-depleted cattle manures temporarily circumvented a low rate of protein fermentation. Results from these swine manure incubations indicate that very different odor formation processes are responsible for the accumulation of odor compounds. In fresh swine manure, protein fermentation was responsible for roughly half of the substrate loss and odor compound production, whereas there was no evidence for protein fermentation in the fresh cattle feedlot manure. Our interpretation of these cattle feedlot and swine manure studies is that starch was the primary (and preferred) substrate for microbial fermentation in cattle feedlot manures; cattle feedlot manure contained high concentrations of fermentable starch relative to protein. Fermentative bacteria in swine manures also utilized starch, but because it was in relatively low abundance, fermentation bacteria also utilized any available protein. Thus, substrate availability likely influenced the diversity and function of the manure microbial community, which in turn control the composition of odorous fermentation products produced from the manure.
|
). Cattle feedlot manures produced numerically higher concentrations of lactate, alcohols, and total VFA per gram of DM than the swine manures. High concentrations of lactate in the cattle manure lowered pH, which would enhance VFA volatility. High concentrations of lactate would also act to limit further production of odor compounds by contributing to VFA toxicity (Switzenbaum et al., 1990). Not only were total VFA concentrations in cattle manure almost twice the content in swine manures, but the composition of individual VFA was quite different. Cattle manure accumulated six times as much butyrate, whereas swine manure accumulated 17 times more branched-chain VFA. Total aromatics in the swine manure were also 50% greater than aromatics in the cattle manure. The differences in butyrate, branched-chain VFA, and aromatic compound content are the likely reasons for differences in perceived odor (dilution to threshold and offensiveness) reported by the general public.
The cattle feedlot and swine manure studies together help describe an emerging picture of the relationships between manure substrate composition, manure microbial communities, and the odor compounds produced during manure fermentation. Clostridium, lactobacillus, and bacillus microorganisms are likely responsible for manure decomposition and odor production (Ouwerkerk and Klieve, 2001; Whitehead and Cotta, 2001) in manure environments. Bacteria in these groups are widely known for their ability to utilize polysaccharides and proteins in anaerobic environments. Furthermore, their ability to form spores gives these microorganisms an additional advantage in environments that experience widely fluctuating moisture, temperature, and oxidative regimes. Our working hypothesis is that differences in substrate availability between cattle and swine manure select for different bacterial species within the Clostridium, lactobacillus, and bacillus groups. The swine manure microbial community appears to be receptive to a wider variety of substrates, whereas the cattle manure microbial community specializes in starch fermentation, which is very abundant, to the exclusion of other substrates. Ultimately, the availability of manure substrates will determine the types of odor compounds produced. Microorganisms, adapted to particular substrate profile, utilize well-defined microbial pathways for starch and protein utilization to produce very predictable, malodorous products (Moat, 1979). Thus, the role of diet and digestive processes in swine and cattle are tightly linked to fresh manure substrate composition, manure microbial communities, and potential odor production. Environmental conditions (chiefly temperature, moisture, and oxygen availability) determine the rate and extent of microbially mediated substrate conversion to odor compounds.
|
Implications |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionImplicationsLiterature Cited |
|---|
|
Footnotes |
|---|
2 The author acknowledges the secretarial assistance of J. Byrkit and the technical assistance of J. Barkhoff, T. Post, and J. Waechter.
3 Correspondence: P.O. Box 166 (phone: 402-762-4208; fax: 402-762-4209; E-mail: miller{at}email.marc.usda.gov).
Received for publication February 19, 2003. Accepted for publication May 5, 2003.
|
Literature Cited |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionImplicationsLiterature Cited |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. L. Conn, E. Topp, and G. Lazarovits Factors Influencing the Concentration of Volatile Fatty Acids, Ammonia, and Other Nutrients in Stored Liquid Pig Manure J. Environ. Qual., January 25, 2007; 36(2): 440 - 447. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. N. Miller and B. L. Woodbury A Solid-Phase Microextraction Chamber Method for Analysis of Manure Volatiles J. Environ. Qual., October 27, 2006; 35(6): 2383 - 2394. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. L. Archibeque, D. N. Miller, H. C. Freetly, and C. L. Ferrell Feeding high-moisture corn instead of dry-rolled corn reduces odorous compound production in manure of finishing beef cattle without decreasing performance J Anim Sci, July 1, 2006; 84(7): 1767 - 1777. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. J. Kerr, C. J. Ziemer, S. L. Trabue, J. D. Crouse, and T. B. Parkin Manure composition of swine as affected by dietary protein and cellulose concentrations J Anim Sci, June 1, 2006; 84(6): 1584 - 1592. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. N. Miller and E. D. Berry Cattle Feedlot Soil Moisture and Manure Content: I. Impacts on Greenhouse Gases, Odor Compounds, Nitrogen Losses, and Dust J. Environ. Qual., March 1, 2005; 34(2): 644 - 655. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
