JEQ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Web of Science (5)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Merrill, L.
Right arrow Articles by Halverson, L. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Merrill, L.
Right arrow Articles by Halverson, L. J.
Agricola
Right arrow Articles by Merrill, L.
Right arrow Articles by Halverson, L. J.
Related Collections
Right arrow Air Pollution
Right arrow Animal Waste
Journal of Environmental Quality 31:2074-2085 (2002)
© 2002 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America

TECHNICAL REPORTS
Waste Management

Seasonal Variation in Microbial Communities and Organic Malodor Indicator Compound Concentrations in Various Types of Swine Manure Storage Systems

L. Merrilla and L. J. Halverson*,b

a Department of Microbiology, 2537 Agronomy Hall, Iowa State University, Ames, IA 50011-1010
b Departments of Agronomy and Microbiology, 2537 Agronomy Hall, Iowa State University, Ames, IA 50011-1010

* Corresponding author (larryh{at}iastate.edu)

Received for publication December 28, 2001.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Anaerobic manure storage systems are one of the major contributors to the odor and environmental pollution associated with swine (Sus scrofa) production systems. The microbial ecology of manure storage systems and the relationships between microbial communities and odor production are largely unknown. In this study, we used community fatty acid methyl ester (FAME) analysis to generate lipid profiles to assess seasonal differences among microbial communities inhabiting various types of outdoor swine manure storage systems. Concurrently, we measured manure concentrations of several malodor indicator compounds as well as pH, temperature, and solids content. Principal components analysis (PCA) showed that there are differences in FAME profiles among the swine manure storage systems examined and most of the variation was in the relative abundance of 18:0, 18:1{omega}7t, 18:1{omega}7c/{omega}9t/{omega}12t, and 16:1{omega}7t/i15:0 2OH FAMEs. The PCA of the FAME profiles revealed that the phototrophic systems were more similar to each other and that the nonphototrophic systems were more similar to each other than they were to phototrophic lagoons. There were seasonal changes in the FAME profiles in the phototrophic systems and the concrete nonphototrophic basin (CNPB), and in one phototrophic system, the FAME profiles more closely resembled a CNPB FAME profile during the winter than the other phototrophic lagoons. In the phototrophic lagoon systems, there was a direct correlation between the abundance of the FAMEs identified in the PCA and manure concentrations of phenol, p-cresol, and 4-ethyl phenol. In the CNPB, there was a negative correlation between the total phenolics concentration and the 18:1{omega}7t FAME. Our results indicate that community FAME profiles could be used as a diagnostic tool for obtaining preliminary evidence that management practices are altering the system's microbial community to one that favors less air pollution potential.

Abbreviations: CNPB, concrete nonphototrophic basin • ENPB, earthen nonphototrophic basin • FAME, fatty acid methyl ester • PCA, principal components analysis • SSPL, single-stage phototrophic lagoon • TS1PL, two-stage first phototrophic lagoon • TS2PL, two-stage second phototrophic lagoon


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
AS swine production systems increase in size, the storage and processing of large amounts of animal manure becomes a serious management and environmental issue (Harkin, 1997). In the USA, there are approximately 157 000 swine production facilities producing more than 118 520 000 Mg of manure annually (American Society of Agricultural Engineers, 1988). In Iowa alone, it is estimated that 14.5 million swine produce 21.3 million Mg of liquid manure per year (J. Lorimor, personal communication, 1998). To handle the large volume of manure produced, most swine confinement operations store the animal manure in deep pits beneath farm buildings or outside in earthen lagoons or concrete basins for periods often exceeding a year before land application (Harkin, 1997; Zahn et al., 1997). These anaerobic storage systems can produce malodorous compounds that can potentially contribute to contamination of the atmosphere, surface water, and ground water near the storage sites, and some of these compounds (phenol and p-cresol) are USEPA priority pollutants (Zahn et al., 1997, 2001b). Within the last 10 years, concerns about odor nuisance and potential health effects associated with manure storage have become a major political and social issue (Harkin, 1997; Marks and Knuffke, 1998), which has resulted in renewed interest in identifying factors that affect malodor emissions from large livestock facilities with the ultimate goal of reducing odor pollution. Because it is primarily microbiological processes that produce and breakdown the malodorous compounds, our poor understanding of the microbial ecology of anaerobic, outdoor swine manure systems has hindered our ability to establish links between microbial communities and malodor emission and develop management practices that reduce odor emissions (Miner, 1995; Zhu and Jacobson, 1999).

Swine manures are comprised of concentrated organic materials, which results in a high oxygen demand for their decomposition. Consequently, the manure becomes anaerobic if not aerated. Decomposition of the organic material can result in the production of malodorous, volatile low molecular weight compounds such as small-chain acids and aromatic compounds such as phenol, indoles, and cresols as well as ammonia and hydrogen sulfide. More than 200 volatile organic and inorganic compounds can be emitted from these systems (Mackie et al., 1998; Zahn et al., 1997). Air and manure concentrations of small-chain acids, phenolics, indoles, and cresols have been shown to be good indicators of odor emission potential and offensiveness (Spoelstra, 1977, 1980; Zahn et al., 2001a,b).

Most studies on the microbiology of odor have focused on the metabolic characteristics or population abundance of species that are presumed to play an important role in odor production or consumption (Kobayashi and Kobayashi, 1995; Zhu and Jacobson, 1999). The primary inoculum of the storage systems are the microorganisms excreted by the animals, and our knowledge of their identity is based primarily on culture-based studies of bacteria present in swine feces or swine digestive tracts (Butine and Leedle, 1989; Rall et al., 1970; Russell, 1979; Salanitro et al., 1977; Zhu and Jacobson, 1999). More recently, 16S rDNA sequence analysis of cultured isolates and cloned 16S rDNA genes from community DNA obtained from gastrointestinal contents (Lin et al., 1997; Pryde et al., 1999; Simpson et al., 1999) and manure (Cotta et al., 1999; Leung and Topp, 2001; Whitehead and Cotta, 1999) has been used to identify organisms that potentially inhabit outdoor manure storage facilities. Microbial populations are not expected to be consistent within and between systems since it has been previously shown that swine intestinal microorganisms vary with animal age (Simpson et al., 1999; Varel et al., 1987) and diet (Moore et al., 1987). Consequently, the microbial composition of the fecal inocula will differ between systems and over time. Due to the open nature of outdoor storage facilities, additional inocula can arise from indigenous organisms present in clay liners, or from microorganisms on airborne particles, vegetation, soil, or animals that fall into the storage facilities. Also, management practices and environmental conditions will influence physical and chemical properties of the storage facilities, which will affect survival of particular species populations. Microbial communities would be expected to differ among various types of systems and seasonally within a system, because of the dynamic nature of the inocula composition, management practices, and seasonal variation in environmental conditions. In fact, changes in microbial community structure are evident in some lagoon systems during the summer months since they develop a visibly deep purple color due to a bloom of purple nonsulfur photosynthetic bacteria. Yet, it is not clear to what extent microbial communities change in these systems or in other storage systems that do not exhibit a phototrophic bloom.

Fatty acid methyl ester (FAME) analyses increasingly are being used to generate fingerprints to characterize microbial communities in natural habitats. Most community studies have isolated the phospholipids (PLFA) of the community prior to generating FAMEs for analyses (Bossio et al., 1998; Carpenter-Boggs et al., 1998; Findlay and Watling, 1998; Frostegard et al., 1997; Guckert et al., 1985; Ibekwe and Kennedy, 1998b), although alternative approaches that do not require isolation of the phospholipids prior to FAME generation have been explored, such as the MIDI method (Cavigelli et al., 1995; Haack et al., 1994; Ibekwe and Kennedy, 1998a; Schutter and Dick, 2000) and an ester-linked method (Ritchie et al., 2000; Schutter and Dick, 2000). In this study the MIDI-FAME technique was used to generate FAME profiles as an indicator of biological changes because it is less labor-intensive than PLFA analysis and provided an opportunity to develop more in-depth seasonal comparisons among and within various storage systems. This study reports evidence on potential differences in manure microbial communities among various types of manure storage systems and that seasonal shifts within communities can occur within a system. Furthermore, we have evidence showing that there is a direct relationship between specific FAMEs and the manure concentrations of various malodor indicator compounds in these systems.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Sample Collection
Liquid manure samples were collected from outside manure storage systems located at various swine confinement operations in central Iowa. The types of systems include a concrete and earthen basin, a phototrophic two-stage lagoon, and a phototrophic single-stage lagoon, and their classification is based on the criteria established by Zahn et al. (2001b). The earthen, two-stage phototrophic lagoon consisted of a primary lagoon (TS1PL), which was 64 x 64 x 7 m (length x width x depth), and a secondary lagoon (TS2PL), which was 61 x 64 x 4 m. This system contained effluent from 1200 sows with piglets and 3000 feeder pigs. The earthen single-stage phototrophic lagoon (SSPL) was 24 x 12 x 3 m and contained effluent from 200 finishing animals. The earthen nonphototrophic basin (ENPB) was 29 x 29 x 3 m, and contained effluent from 1200 finishing animals. The concrete nonphototrophic basin (CNPB) was 24.4 m in diameter x 2.44 m deep, and contained effluent from approximately 1200 finishing animals. At all of the farms, the animals live on slotted floors inside confinement buildings and their manure falls into shallow pits beneath the buildings before being emptied into outside storage systems through a pipe. A single lagoon receives all the manure from a building while in a two-stage system, the first lagoon receives manure from the building(s) and is connected by a pipe to a second lagoon. In our study, the SSPL, TS1PL, and TS2PL systems experienced a phototrophic bloom, and the manure was a pink–purple color in each of these lagoons from June until October (Do et al., 1998). The ENPB and CNPB basin systems did not turn visibly purple during our study.

Sampling from all of the storage systems, except ENPB, occurred between March 1997 and November 1998 while ENPB was sampled between November 1997 and November 1998. We collected samples monthly during December and January, biweekly from August through November and February through May, and weekly during June and July. From December 1997 through February 1998 all of the earthen lagoons had a layer of ice approximately 0.3 m thick, and samples were collected from underneath the ice layer. In the lagoon systems, samples were collected from the top 30 cm of the manure at random points approximately 3 m from the edge of the lagoon. Previous work (DiSpirito et al., 1995) showed that surface samples would represent the entire lagoon above the sludge layer due to the high mixing rates of lagoons during the summer; we did not assess mixing rates during winter. Samples were collected from the top 30 cm of manure along the outside edge of the CNPB system. Three 250-mL samples were collected from the TS1PL and TS2PL systems and two 250-mL samples were collected from the SSPL, CNPB, and ENPB systems. All samples were kept on ice and processed within 6 h of collection.

Physical Parameters
Air and manure temperatures were recorded on site at the time of sampling. The sample pH was measured in the laboratory after the samples had equilibrated to room temperature. Aliquots of the manure (2.5 to 5.0 mL) were used to measure total and volatile solids in samples from January through November 1998 with standard protocols (American Public Health Association, 1992).

Fatty Acid Methyl Ester Analysis
Sixty milliliters of manure were centrifuged for 10 min at 12 298 x g (Marathon 21K centrifuge; Fisher Scientific, Pittsburgh, PA). Following centrifugation, two subsamples of pelleted material were collected with a sterile metal spatula and transferred to 13- x 100-mm screw top culture tubes and stored at -20°C until FAME analysis. Within two weeks of sampling, the manure samples were thawed at room temperature, and the FAMEs were extracted with modifications of the Microbial Identification System anaerobic extraction protocol (Microbial ID, 1996). For the phototrophic systems, 1.25 mL of methyl-tert butyl ether (MTBE) and hexane was used for extraction of the FAMEs. Due to the abundant cell biomass in the CNPB and ENPB systems, we used 2.5 mL of MTBE and hexane to extract and dilute the sample to obtain a total FAME area that was comparable with those for the phototrophic systems. Samples were stored at 4°C for up to 2 wk.

The FAME samples were analyzed on a Hewlett–Packard (Rolling Meadows, IL) 6890 gas chromatograph equipped with an HP Ultra2 capillary column (crosslinked 5% Ph Me silicone; 25-m length x 0.2-mm i.d. x 0.33-mm film thickness) and a flame ionization detector. The majority of samples were analyzed with the MIDI aerobe method (Microbial ID, 1996), which separates FAMEs from 9 to 20 carbons in length. A subset of extracted samples was also analyzed for the presence of FAMEs that are typically associated with eukaryotic cells by using the MIDI eukarya method, which separates peaks up to 30 carbons in length. We present FAME data in units of percent of the total area of named FAMEs using standard fatty acid nomenclature. The length of the fatty acid and amount of saturation are reported with the number of carbon atoms followed by a colon and the number of double bonds after the colon. Position of the double bond is counted from the methyl end ({omega}) of the molecule and can be either in a cis (c) or trans (t) orientation. Other notations are Me for methyl group, OH for hydroxy, and the prefixes i and a for iso- and anteiso-branched FAMEs. It should be noted that the 18:1{omega}7c, 18:1{omega}9t, and 18:1{omega}12t FAMEs are summed and reported together because they have similar retention times using the MIDI system. Similarly, 16:1{omega}7c and i15:0 2OH cannot be reliably separated and are reported together. Because nonmicrobial compounds from the manure sample could elute at retention times similar to phospholipid or lipopolysaccharide FAMEs, MIDI peak identifications should be considered preliminary until the structures of each peak are verified by gas chromatography–mass spectrometry (GC–MS) analyses.

Measurement of Malodorous Organic Indicator Compounds
From October 1997 through November 1998, the same samples that were used for FAME analysis were also used for analysis of phenolic compounds. After centrifugation for 10 min at 15 770 x g at room temperature, the supernatant was decanted and combined for each site in glass screw-top bottles and stored at 4°C for no longer than 24 h. Phenolics from the manure were concentrated by solid-phase extraction (SPE) with 6-mL tubes packed with 0.5 g ENVI Chrom-P (Supelco, Bellefonte, PA) that were conditioned according to manufacturer's instructions. Prior to extraction, samples were filtered through a Whatman (Maidstone, UK) GF/B filter (1.2-µm particle retention) to remove solids. For the basins and phototrophic lagoons, duplicate 5- to 10-mL and 50- to 100-mL amounts of supernatant were analyzed, respectively. After sample application, the solid-phase extraction packing material was dried under vacuum for a maximum of 2 min to remove water and any dissolved aqueous material with a vacuum manifold system (Alltech Associates, Deerfield, IL). Then, the solid-phase extraction resin was soaked for 1 min with 2 mL of MTBE to extract the phenolics from the resin and an additional 2 mL of MTBE was added to elute the phenolics under vacuum. MTBE was continually added until 5 mL of sample was collected. For all samples, approximately the first 1 mL of material collected was a brown layer that separated distinctly from the upper clear layer. This bottom layer was aqueous material adsorbed to the packing material that was not removed by the drying step. The upper clear layer was transferred into autosampler vials for analysis by gas chromatography within 24 h.

Phenols and indoles were analyzed with a HP 6890 fitted with a HP Ultra2 capillary column (crosslinked 5% Ph Me silicone, 25-m length x 0.2-mm i.d. x 0.33-mm film thickness) and a flame ionization detector. The oven temperature was increased from 65 to 185°C at a rate of 10°C min-1, maintained at 185°C for 1 min, and then the temperature was increased at a rate of 20°C min-1 to a final temperature of 275°C. The final temperature was maintained for 5 min. The hydrogen carrier gas was held constant at a flow rate of 1.2 mL min-1. The standards mixture consisted of dilutions of a 500 mg L-1 phenol and p-cresol and 250 mg L-1 4-ethyl phenol, indole and 3-methyl indole solution. Recovery of the standard mix from filtered manure samples spiked with a dilution of the standards mixture ranged from 82 to 104% for the individual analytes in the mixture with a mean recovery for all the individual analytes of 95 ± 7% (mean ± standard deviation, n = 12).

Initial analysis of samples on a HP 6890 gas chromatograph with a HP 5972 mass spectrophotometer detector was performed to identify the extracted peaks. Samples were separated with the same column, temperature program, and flow rates as above but with helium as the carrier gas. The mass spectrophotometer transfer line temperature was held at a constant 260°C, the electron multiplier voltage was 1700 V, and the resulting ions were scanned from 10 to 550 atomic mass units every 1.5 s. Resulting spectra were compared with the National Institute of Standards and Technology library package provided by Hewlett–Packard, and the following compounds were identified with greater than 90% matching: phenol, 4-methyl phenol (p-cresol), 2-ethoxy phenol, 4-ethyl phenol, piperidinone, 1,2-benzenediol, indole, 3-methyl indole, and 1,3-2H-indole-2-one.

Statistical Analysis
To analyze the FAME profiles, we used principal components analysis (PCA) performed with the PRINCOMP command in SAS (SAS Institute, 1996) with the covariance matrix. To facilitate visualization of differences in the various systems over the two-year duration of our study, we performed separate PCA of the FAME profiles for all the systems during the spring (March to May), in June, in July, in the fall (August to October), and in the winter (November to February). To compare patterns seen in plots of the PCA at different times of the year, we used data matrices composed of the same FAMEs and standardized axes length. When a FAME was not detected in a system but was present in other systems that particular FAME was given a zero value in the profiles where it was not detected. Pearson correlation coefficients and probabilities were determined by multivariate pairwise correlation with SAS.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Physical Characteristics of Storage Systems
Surface manure temperatures were similar among the manure systems throughout the year (Fig. 1) and were slightly warmer during the second year than the first year. From December to early April, the surface layers of the lagoons were frozen whereas the basins stayed in an unfrozen, slushy state. Total and volatile solids content of the manure increased after pumping and stirring events, but otherwise were relatively constant throughout the year. The amount of total solids in the phototrophic lagoons was approximately one-third or less of that in the basin systems. Volatile solids in the phototrophic systems were 40 to 50% of the total solids content (Table 1). There was no statistically significant correlation between manure temperature and total or volatile solids content, except in the SSPL system where there was a negative correlation between temperature and total (r = -0.557; p = 0.014) and volatile (r = -0.546; p = 0.019) solids content. From July through November, when the phototrophic systems experienced a phototrophic bloom, the manure pH for each system was consistently above that system's average pH. There was no significant correlation between manure temperature and pH in the systems, except in TS2PL, which exhibited a positive correlation (r = 0.621; p = 0.002).



View larger version (27K):
[in this window]
[in a new window]
  		Fig. 1. Manure temperature of anaerobic swine manure storage systems.

 

View this table:
[in this window]
[in a new window]
  		Table 1. Physical parameters of swine manure storage systems.

 
Fatty Acid Methyl Ester Analysis
The FAME analysis procedure did not include fatty acids in the dissolved organic matter present in the supernatant, and the pelleted manure sample contained plant and animal cells and insoluble organic matter in addition to microbial cells. We assessed the contribution of nonmicrobial sources of lipids to the FAME profiles through two approaches; filtration to remove cells and material greater than 6 µm and assessment of the contribution of long chain and polyunsaturated FAMEs to the profiles. There was no significant difference in the number, type, or relative amount of FAMEs in the profiles between the filtered and nonfiltered samples (data not shown). Based on MIDI-FAME analysis with the eukarya and aerobe method, we detected the polyunsaturated FAME 18:3 {omega}6c(6,9,12) in 31 out of 116 CNPB samples, which comprised <2% of the total FAMEs, 20:4{omega}6,9,12,15c in 3 out of 208 TS1PL samples, which comprised <1% of the total FAMEs, and 20:3 {omega}6,9,1c in 1 out of 208 TS1PL samples, which comprised <1% of the total FAMEs. The number of FAMEs with >20 carbons contributed less than 1 to 5% to the FAME profiles, and the CNPB contained the greatest number and amounts of FAMEs greater than 20 carbons. Taken together, these results suggest that the contribution of FAMEs from eukaryotic cells to the FAME profiles was small.

In general, there were 16 to 21 FAMEs detected in each analysis, and the number of FAMEs detected in the manure obtained at different times of the year or different years did not change significantly for a particular system. Among all manure storage systems examined only 45 FAMEs were detected in 10% or more of the samples. Only six FAMEs were detected in 95% of the samples of any particular manure storage system (Table 2). Although most FAMEs were detected in all the systems, there were also system-specific FAMEs (Table 2). The large number of branched FAMEs that are typically found in prokaryotic cells (Harwood and Russell, 1984) suggests that a significant number of the FAMEs were probably derived from microbial cells.


View this table:
[in this window]
[in a new window]
  		Table 2. Fatty acid methyl esters (FAMEs) most commonly found in swine manure storage systems.{dagger}

 
Fatty Acid Methyl Ester Comparisons among Manure Systems
We show representative PCA of the FAME profiles to illustrate our key findings and additional PCA analyses that support our conclusions are available elsewhere (Merrill, 1999). In general, the FAME profiles of the basins and phototrophic systems were different from each other during the summer (Fig. 2) and fall months (data not shown). For PCAs from June through October, PC1 accounted for more than 80% of the total variation and PC2 contributed relatively little (2–9%) to the discrimination between samples. PC1 of the July 1998 FAME profiles (Fig. 2) separates the basins from the phototrophic lagoons while PC2 separates the profiles of the two basin systems from each other. Similar patterns were observed in PCA of samples collected during June and July 1997 and June 1998 (data not shown).



View larger version (27K):
[in this window]
[in a new window]
  		Fig. 2. Principal components analysis (PCA) plot of various swine manure storage system fatty acid methyl ester (FAME) profiles during July 1998. Values in parentheses indicate the contribution of PC1 or PC2 to the total variation in the FAME profiles. Circles were drawn to facilitate elucidation of the patterns. Two-stage first phototrophic lagoon (TS1PL) ({circ}); two-stage second phototrophic lagoon (TS2PL) ({square}); single-stage phototrophic lagoon (SSPL) ({triangleup}); earthen nonphototrophic basin (ENPB) ({diamond}); concrete nonphototrophic basin (CNPB) ({triangledown}).

 
During the winter, FAME profiles of the phototrophic lagoons were more similar to the basin systems with PC1 and PC2 accounting for 70% and 16% of the sample variation, respectively (Fig. 3) . A similar pattern was observed in spring 1998 (data not shown). PC1 of the November 1997 to February 1998 samples separates the early winter (November and December) samples from the phototrophic lagoons from the late winter (January and February) samples of the phototrophic systems and all of the samples from basins (Fig. 3). In November, all of the samples from the phototrophic lagoons had positive PC1 scores, and the value of these scores decreased through the winter. By February, the FAME profiles of the phototrophic systems became more similar to the FAME profiles of the basins. During this period there was little change in the FAME profiles of the basins. PC2 primarily provided differentiation between the CNPB and ENPB. The FAME profiles of TS1PL, SSPL, and the basin systems did not change significantly until the beginning of summer.



View larger version (28K):
[in this window]
[in a new window]
  		Fig. 3. Principal components analysis (PCA) plot of various swine manure storage system fatty acid methyl ester (FAME) profiles during November 1997 through February 1998. Values in parentheses indicate the contribution of PC1 or PC2 to the total variation in the FAME profiles. Two-stage first phototrophic lagoon (TS1PL) ({circ}); two-stage second phototrophic lagoon (TS2PL) ({square}); single-stage phototrophic lagoon (SSPL) ({triangleup}); earthen nonphototrophic basin (ENPB) ({diamond}); concrete nonphototrophic basin (CNPB) ({triangledown}).

 
Regardless of the season examined, the same set of FAMEs (16:1{omega}7c/i15:0 20H, 18:1{omega}7c/{omega}9t/{omega}12t, 18:1{omega}7t, 18:0, and 16:0) (Table 3) had high eigenvectors for PC1. In the basins, we never detected the 18:1{omega}7c/{omega}9t/{omega}12t FAME, and the 16:1{omega}7c/i15:0 20H FAME appeared infrequently and at relatively low percent of the total FAMEs, whereas 18:0 and 18:1{omega}7t accounted for 30 to 40% of the total profile. In the ENPB, 16:0 accounted for approximately 30% of the total profile. In the CNPB, there was some seasonal variation in the relative amount of 16:0 with greater proportions in the fall and early winter months. In the phototrophic systems, the 16:1{omega}7c/i15:0 20H, 18:1{omega}7c/{omega}9t/{omega}12t, and 18:1{omega}7t FAMEs exhibited seasonal trends while the 16:0 comprised a relatively constant proportion of the total FAMEs. In the phototrophic lagoons, both the 16:1{omega}7c/i15:0 20H and 18:1{omega}7c/{omega}9t/{omega}12t FAMEs accounted for 40 to 50% of the profiles during the summer and less than 10% of the profiles during winter and spring. The FAMEs 18:0 and 18:1{omega}7t were at their lowest relative percent during the summer (5% and undetected, respectively) and peaked during the late winter with relative proportions of 30% for 18:0 and 10 to 25% for 18:1{omega}7t.


View this table:
[in this window]
[in a new window]
  		Table 3. Eigenvectors for fatty acid methyl esters (FAMEs) that contribute most to the differentiation among the swine manure storage facilities.{dagger}

 
Seasonal Variation of Fatty Acid Methyl Ester Profiles within Manure Systems
For all of the phototrophic systems, there was a temporal shift within PC1 and PC2 added little to the variation in the FAME profiles. We show representative PCA plots to illustrate the seasonal changes in the FAME profiles observed in the phototrophic lagoons (SSPL, Fig. 4) and the CNPB (Fig. 5) . There was no detectable seasonal pattern in the FAME profiles obtained from the ENPB.



View larger version (31K):
[in this window]
[in a new window]
  		Fig. 4. Seasonal principal components analysis (PCA) plot of the single-stage phototrophic lagoon (SSPL) fatty acid methyl ester (FAME) profiles. Values in parentheses indicate the contribution of PC1 or PC2 to the total variation in the FAME profiles. Spring is March–May, Summer is June–July, Fall is August–October, and Winter is November–February.

 


View larger version (25K):
[in this window]
[in a new window]
  		Fig. 5. Seasonal principal components analysis (PCA) plot of the concrete nonphototrophic basin (CNPB) fatty acid methyl ester (FAME) profiles. Values in parentheses indicate the contribution of PC1 or PC2 to the total variation in the FAME profiles. Summer is June–July, Fall is August–October, and Winter is November–February.

 
In the SSPL, PC1 separated the winter through spring FAME profiles from those of summer through fall (Fig. 4). The FAMEs that contribute the most to the variation in PC1 (eigenvector values in parentheses) were 18:1{omega}7t (-0.42), 18:0 (-0.54), 16:1{omega}7c/i15:0 20H (0.43), and 18:1{omega}7c/{omega}9t/{omega}12t (0.57). These are the same FAMEs that accounted for the variation seen in comparisons among the various systems (Table 3). The summer samples contained larger relative amounts of the 16:1{omega}7c/i15:0 20H and 18:1{omega}7c/{omega}9t/{omega}12t FAMEs and lower relative amounts of 18:1{omega}7t and 18:0 compared with the samples collected from January through May.

Principal components analysis of FAME profiles from the CNPB shows some seasonal variation in the FAME profiles, although relative to the phototrophic systems, the magnitude of this variation is lower (Fig. 5). For both years, the spring samples had negative PC1 scores whereas the July through December samples had positive PC1 scores. The June 1997 samples were most similar to the spring samples of both years whereas the June 1998 samples were most similar to the summer, fall, and winter samples. The FAMEs that contributed the most to the variation in PC1 (eigenvector values in parentheses) were 18:0 (0.82), 18:1{omega}7t (-0.40), 16:0 (0.24), 18:1{omega} 6,9c/a18:0 (-0.20), 18:1{omega}9c (-0.16), and 20:0 (-0.16). During both years of our study, the spring FAME profiles had lower proportions of 18:0 and 16:0 and greater proportions of 18:1{omega}7t, 18:1{omega}6,9c/a18:0, 18:1{omega}9c, and 20:0 FAMEs compared with the profiles for the rest of the year. Although the FAMEs 18:0 and 18:1{omega}7t explained temporal variability in both the phototrophic lagoons and the CNPB systems, in the CNPB the two were negatively correlated whereas in the phototrophic systems they were positively correlated. Contents of the CNPB system were removed in March 1997, and in January and November 1998.

Fatty Acid Methyl Ester Profiles of Purple, Nonsulfur Phototroph Isolates
The MIDI-FAME profiles of nine isolates of purple, nonsulfur phototrophs obtained from the TS1PL and TS2PL systems were identified as members of the Rhodobacter genus. The isolation and characterization of these Rhodobacter isolates is described elsewhere (Do, 2001). The isolates were comprised primarily of the 18:1{omega}7c/{omega}9t/{omega}12t (92.1 ± 1.8%; mean ± standard error of the mean; n = 9), 18:0 (2.4 ± 0.25%), and 16:1{omega}7t/i15:0 20H (1.6 ± 0.1) FAMEs.

Manure Concentrations of Malodor Indicator Compounds
During the second year of our study, we examined the manure concentrations of various malodor indicator compounds. In the phototrophic lagoon, there was an increase in the amount of phenol, p-cresol, 4-ethyl phenol, and indoles (summation of indole and 3-methyl indole) during the winter months, followed by a decrease to low or nondetectable levels in June and July (Fig. 6) . The seasonal trend was similar for SSPL (data not shown). There was a clear sequential spike in the concentrations of these compounds in early June that appeared first in TS1PL and then in TS2PL. p-Cresol comprised approximately 80% of the phenolics measured in the slurries of the phototrophic systems.



View larger version (24K):
[in this window]
[in a new window]
  		Fig. 6. Seasonal changes in malodor indicator compounds in manure slurries of the two-stage phototrophic lagoon system. (A) p-Cresol (2-methyl phenol) ({blacktriangledown},{triangledown}) and 4-ethyl phenol (•,{circ}). (B) Phenol ({diamondsuit},{diamond}) and indoles (indole + 3-methyl indole {blacksquare},{square}). Open symbols, two-stage first phototrophic lagoon (TS1PL); solid symbols, two-stage second phototrophic lagoon (TS2PL). Values are the averages of two samples for each lagoon, and the coefficient of variation over all sampling times was about 5%. A similar profile was observed for the single-stage phototrophic lagoon (SSPL) system (data not shown).

 
In the CNPB, phenol and p-cresol comprised >85% of the total amount of phenolics measured and there was little seasonal variation in the amounts of the other malodor indicator compounds (Fig. 7A) . There was an increase in the amount of phenol that coincided with a decrease in the amount of p-cresol in the manure in the months prior to pumping out events in late January and early November to remove manure from the system (Fig. 7A). Following manure removal in early November 1998, there was a sharp decrease in the amount of phenol measured in the manure (Fig. 7A). In the ENPB, phenol and p-cresol generally comprised >80% of the indicator compounds measured (Fig. 7B), and there was a regular periodic increase in the amount of p-cresol and to a lesser extent phenol (Fig. 7B). The periodic increase in p-cresol did not correspond to the periodic input of manure into the lagoon, which occurred every other week. On average there was three times more phenol in samples from the CNPB than the ENPB and there were comparable amounts of p-cresol in both systems (Fig. 7)



View larger version (33K):
[in this window]
[in a new window]
  		Fig. 7. Seasonal changes in malodor indicator compounds in the (A) concrete nonphototrophic basin (CNPB) and (B) earthen nonphototrophic basin manure slurries. The arrows indicate when the system contents were removed. We were unable to obtain samples in the CNPB system between February and April after the contents had been removed.

 
Phenolic Concentrations and Fatty Acid Methyl Ester Profile Relationships
In the phototrophic lagoons, there is a strong correlation between the relative abundance of four FAMEs identified from PCA and the concentration of malodor indicator compounds detected in the manure (Fig. 8 and Table 4). The measurement of detectable concentrations of phenolics (phenol, p-cresol, 4-ethyl phenol) positively correlated with the relative amounts of 18:1{omega}7t in the FAME profiles in the winter (Fig. 8 and Table 4). In the spring, the proportion of 18:1{omega}7t and the total phenolic concentration decreased to near undetectable levels (Fig. 8). A similar trend also was observed for 18:0, with its relative amount decreasing from 30 to 5% of measured FAMEs during the late spring, which was positively correlated with the total phenolic concentration (Table 4). Conversely, the 18:1{omega}7c/{omega}9t/{omega}12t FAME was negatively correlated with the amount of phenols and indoles (Fig. 8 and Table 4). For all the phototrophic systems during 1998, the relative proportion of the 18:1{omega}7c/{omega}9t/{omega}12t FAME was the highest when the phenolic concentrations were the lowest (Fig. 8 and Table 4). Similarly, the 16:1{omega}7t/i15:0 20H FAME exhibited a negative correlation with the concentration of phenols and indoles in the manure (Table 4). There was no significant correlation between total and volatile solids content and total phenolic concentration in any of the phototrophic lagoons (Table 4). Furthermore, there was a negative correlation between pH and total phenolic concentration in TS2PL and SSPL. There was a significant negative correlation between surface manure temperature and total phenolic concentration for all the phototrophic systems (Table 4).



View larger version (29K):
[in this window]
[in a new window]
  		Fig. 8. Seasonal relationship between total phenolic concentration and proportion of 18:1{omega}7c/{omega}9t/{omega}12t and 18:1{omega}7t fatty acid methyl esters (FAMEs) in the two-stage first phototrophic lagoon (TS1PL). Total phenolics is the summation of phenol, p-cresol, 4-ethyl phenol, and the indoles. 18:1{omega}7t ({blacktriangledown}), 18:1{omega}7c/{omega}9t/{omega}12t (•), and total phenolic concentration ({square}). Similar profiles were observed for the two-stage second phototrophic lagoon (TS2PL) and single-stage phototrophic lagoon (SSPL).

 

View this table:
[in this window]
[in a new window]
  		Table 4. Pairwise correlations between total phenolic concentration and selected fatty acid methyl esters (FAMEs), solids content, temperature, and pH.{dagger}

 
Only the 18:1{omega}7t FAME exhibited a significant correlation with the total phenolic concentration in the CNPB and ENPB systems, although the correlation was negative for CNPB and positive for ENPB (Table 4). For the CNPB system, there was a negative correlation between the total and volatile solids content and total phenolic concentration, although not statistically significant (Table 4), while there was no consistent pattern between volatile and total solids content and total phenolic content in the ENPB system. There was no significant correlation between total phenolic concentration and surface temperature. There was a significant correlation between pH and total phenolic concentration, although the correlation was positive for CNPB and negative for ENPB.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
The FAME analysis showed that there were differences in the FAME profiles among various types of manure storage systems as well as seasonally within a system. We used PCA analysis to reduce the dimensionality of the data and graphically show the separation of the manure systems into clusters, and these analyses clearly illustrate the seasonal differences and similarities among these systems (Fig. 2–5). In general, the systems seem to be more different from each other during the summer and fall months and more similar to each other during the winter and spring. This is in part due to the dramatic change in the FAME profiles of the phototrophic systems in the summer when they are experiencing phototrophic blooms. In the winter the SSPL and TS1PL FAME profiles become more similar to those of the CNPB (Fig. 3). Because the total solids content of the phototrophic systems in the winter months did not approach the amounts measured in the CNPB system (Table 2), it is unlikely that the similarity in FAME profiles during the winter can be explained by reduced microbial activity during concomitant input of fresh manure. In the summer the various storage systems are more different from each other (Fig. 2), but even in the winter when the FAME profiles among the systems are more similar, there appears to be a clear clustering of each system (Fig. 3). The greater similarity in FAME profiles during the winter may be a result of the fact that all of these systems receive manure from swine and that manures from farrowing and feeder operations and finishing operations are not significantly different, a result of reduced mixing of system contents during colder months, or that environmental factors such as temperature strongly influence microbial community structure.

Principal components analysis indicates that a small subset of FAMEs seems to play a significant role in discrimination of FAME profiles among these storage systems. Regardless of the sampling time, 18:1{omega}7t, 18:0, 16:0, 16:1{omega}7c/i15:0 2OH, and 18:1{omega}7c/{omega}9t/{omega}12t were important contributors to the differences among the FAME profiles of phototrophic lagoons and basin systems. In general, none of the five FAMEs that account for most of the variability in PCA of the FAMEs were helpful in inferring the taxonomic composition of the microbial communities due to the ubiquitous distribution of these FAMEs in a wide variety of bacterial species and taxonomic groups. However, it is tempting to hypothesize that the increase in the 18:1{omega}7c/{omega}9t/{omega}12t FAMEs in the phototrophic systems in the summer was due to the bloom of purple, nonsulfur phototrophs since the 18:1{omega}7c/{omega}9t/{omega}12t FAME comprised 92% of the FAME profile of these isolates. The increase in this FAME also coincided with an increase in bacteriochlorophyll content in the manure (Do et al., 1998).

The seasonal change in FAME profiles in the CNPB occurred in the summer (July and August) during both years of our study (Fig. 5 and data not shown), although those changes were not as dramatic as those observed for the phototrophic lagoons. The change in FAME profiles occurred during the period of peak system temperature (Fig. 1), which indicates that temperature may be a key factor influencing the change in FAME profiles. It is unlikely that the seasonal changes are a direct consequence of manure removal from the storage system since that generally occurred during late winter and early spring (February or March), whereas the changes in the FAME profiles occurred in July and August. The ENPB did not exhibit any seasonal change in its FAME profiles that we could detect with PCA, which suggests that either the microbial communities did not change or that lipid profiling did not have the sensitivity to detect these changes.

In the phototrophic systems, there was a statistically significant correlation (p < 0.001) between manure temperature and the FAMEs 18:1 {omega}7t, 18:0, 18:1{omega}7c/{omega}9t/{omega}12t, and 16:1{omega}7c/i15:0 2OH (data not shown). It is conceivable that seasonal changes in FAME profiles were the result of environmental effects on FAME composition of the microbial community, although the saturated to unsaturated and iso to anteiso FAME ratios (Merrill, 1999) are the opposite of what we would anticipate given the well-documented responses of pure bacterial cultures to changes in environmental temperature (Haack et al., 1994; Harwood and Russell, 1984; Kaneda, 1991; Petersen and Klug, 1994; Rose, 1989). This suggests that seasonal changes in FAME profiles in the phototrophic lagoons were not due to temperature-mediated adaptations by microorganisms to environmental temperature. There was no consistent trend between manure temperature and these FAMEs in the CNPB and ENPB systems (data not shown). Fatty acid composition also changes in response to environmental pH (Rose, 1989), but the seasonal variation in pH was generally less than one-half of a pH unit, which probably is too small to cause the dramatic changes in FAME profiles during the summer.

There was great seasonal and system-specific variation in the concentration of malodor indicator compounds in the manure (Fig. 6 and 7). The phototrophic systems had the lowest amounts of malodor indicator compounds, and they exhibited the greatest seasonal change in concentration. It is unlikely that the reduction in phenol and p-cresol in TS1PL and TS2PL was due to the emptying of system contents in early May 1998 because there had already been a significant reduction in the phenol and p-cresol concentration in April (Fig. 6). It is worthwhile to note that the spike in the phenol concentration in June was first detected in the TS1PL and then two weeks later in the TS2PL (Fig. 6). It is unclear whether the spike was due to a change in management practices that resulted in manure entering into the system that had a higher phenol concentration, to a change in manure management practices that altered microbial production or consumption of phenol, or to a change in environmental factors that resulted in a burst in microbial activity. We also measured the concentration of various short-chain acids (C3–C8), and we were unable to detect them in the manure in May through August 1998 (data not shown); these acids were detected in all of the basin system samples. Although it is conceivable that the decrease in manure phenolic concentration during the summer months could be due to a greater flux of such compounds into the atmosphere due to the higher summer temperatures, we do not think it is likely because similar decreases in phenolics were obtained in air samples collected over the TS1PL and TS2PL systems during the summer of 1997 (Do et al., 1998; Zahn et al., 2001b).

In contrast to the lagoons, the basin systems had substantially higher concentrations of malodor indicator compounds (compare Fig. 6 and 7). The CNPB consistently had higher phenolics concentrations and twice the amount of phenol as the ENPB. In the CNPB, the concentrations of indoles and 4-ethyl phenol were relatively constant throughout the year; however, in the ENPB there were periodic increases in their concentration that coincided with larger increases in phenol and p-cresol concentrations. We have no sound explanation for the regular periodic and dramatic increase in the concentration of p-cresol in the ENPB (Fig. 7B), but it is possible that they are due to changes in management practices that resulted in the input of manure into the basin that had a higher p-cresol content or alternatively some management practice affected the production or consumption of p-cresol by the resident microbiota in the basin. The amount of these compounds in the manure contributes directly to the offensiveness or intensity of the malodor. However, since we were unable to reliably quantify the short chain volatile organic acids we were unable to use the recently published odor intensity perception model to assess changes in odor intensity (Zahn et al., 1997, 2001b). It has been recently shown that relatively high concentrations of odorants were more important in predicting odor intensity than the diversity of odorants detected (Zahn et al., 2001a). This would suggest that the higher concentrations of the volatile organic compounds we measured in the basin systems could be used to predict potential odor intensity of these storage systems, although it is often difficult to correlate manure concentration of odorants to the air concentration of the same compounds (Zahn et al., 1997).

For the phototrophic systems, there were statistically significant correlations (Table 4) among the manure concentration of phenolics and the relative abundance of four different FAMEs (18:1 {omega}7t, 18:0, 18:1{omega}7c/{omega}9t/{omega}12t, and 16:1{omega}7c/i15:0 2OH). In the CNPB and ENPB systems there was a statistically significant correlation between the total phenolic concentration and 18:1 {omega}7t FAME (Table 4), although the correlation was positive in the ENPB and negative in the CNPB. These results suggest that different factors influence the abundance of 18:1{omega}7t FAMEs in these different types of systems. The presence of trans monounsaturated FAMEs has frequently been used as an indicator of organic pollutant exposure (Heipieper et al., 1992) and/or nutrient deprivation (Guckert et al., 1986; Kieft et al., 1994). For the phototrophic systems and the ENPB, it is possible that the decrease in 18:1{omega}7t reflects the decrease in the manure phenolics concentration. Alternatively, the abundance of 18:1{omega}7t during the winter and spring months could reflect changes in community composition where the 18:1{omega}7t synthesizing bacteria are in greater abundance or starved. This, however, does not explain the negative correlation between 18:1{omega}7t and the total phenolics concentration in the CNPB.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
In summary, our results show that (i) FAME analysis can be used as a tool to assess the degree to which manure storage system microbial communities are similar or different to one another, (ii) there are dramatic seasonal changes in microbial communities in the phototrophic lagoon systems as well as to a lesser extent changes in the microbial communities in the CNPB system in the summer, (iii) the seasonal changes in the microbial communities are most pronounced during the summer when system contents are warmer, (iv) there are dynamic seasonal changes in the concentration of various volatile organic compounds in swine manure storage systems, and (v) there is a direct correlation between the concentration of phenolic compounds in the manure and the relative abundance of the FAMEs identified in the PCA analysis that contribute to the separation of the FAME profiles among and seasonally within different systems. This work provides a foundation on which to better understand the relationships between microbial community structure and malodor production in these manure storage systems. Ultimately, this will facilitate a better understanding of factors influencing odor emission, better predictive tools for potential odor emission, and better management strategies that favor the establishment of microbial communities that have less odor producing potential.


   ACKNOWLEDGMENTS
 
We thank Naomi Bremer and Janelle McKinney for technical assistance. We thank Dr. Ken Koehler for consultation and discussion on the use and interpretation of PCA and Dr. Jim Zahn for critically reading an earlier version of the manuscript. We thank Y.S. Do for providing the purple, nonsulfur phototrophs isolates for FAME analysis. This project was funded by the Iowa Swine Odor and Waste Management Research Program, the Iowa Academy of Sciences (ISF-98-13), the Hatch Act, and the State of Iowa. Journal Paper no. J-19328 of the Iowa Agriculture and Home Economics Experiment Station, Ames, Iowa, Project no. IOW03439.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


Home page
 J. Environ. Qual.Home page
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]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Web of Science (5)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Merrill, L.
Right arrow Articles by Halverson, L. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Merrill, L.
Right arrow Articles by Halverson, L. J.
Agricola
Right arrow Articles by Merrill, L.
Right arrow Articles by Halverson, L. J.
Related Collections
Right arrow Air Pollution
Right arrow Animal Waste

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Soil Science Society of America Journal Journal of Plant Registrations The Plant Genome