1. Introduction
The biodegradation of organic waste is one of the important steps of the carbon cycle in nature. As one of the biodegradation technologies, composting is often used to transform agricultural organic waste into stable usable resources under controlled conditions. It is an aerobic and spontaneous heating process of recycling organic waste which is controlled by microbial communities. Composting has the advantages of being low cost, environmentally friendly and it produces no further pollution [1,2]. As an important basis of agriculture waste, lignocellulose is one of the important sources of renewable carbon on Earth [3]. It is also known as a persistent organic compound, because of its complex cross-linked structure and stable composition which forms a matrix accompanying cellulose. The presence of the matrix strongly impedes the microbial depolymerization of cellulose and hemicellulose, so the degradation of lignin has become a difficult process when lignocellulose degrades in agricultural waste composting [4,5]. It is known that the decomposition of organic substrates in composting is mainly carried out by microorganisms. However, Actinomycetes, considered a specific group of Gram-positive bacteria that can form branching hyphae, make a great contribution to the decomposition of complex organic compounds like lignin [5] and other organic matters [6] in composting. In addition, Actinomycetes also plays a significant role in the bio-decomposition of agricultural waste during composting [7].
Some studies have explored the interaction between changes in environmental factors and microbial communities during composting [8,9,10,11,12]. Sundberg [11] reported that different pH settings had a significantly different influence on microbial communities during the composting process. Liang [9] drew the conclusion that the effect of pile temperature on the metabolism of the microbe was less than that of moisture content. A study by Tang [12] showed that water soluble carbon (WSC) and pile temperature were dominant factors compared with other parameters. These results are conflicting and only a few physico-chemical parameters were measured separately. Therefore, these reports cannot provide enough comprehensive and systematic understanding of the composting of agricultural waste. Zhang [2] separated out and prioritized the relative influence of several environmental factors on the community change of bacterial and fungi by evaluating them simultaneously. However, the influences of physico-chemical parameters on Actinomycetes communities during agricultural waste composting are still not clear.
Denaturing gradient gel electrophoresis of polymerase chain reaction (PCR-DGGE), as a widely applied molecular biological technology, has been used in composting studies. This technology was frequently used to generate a fingerprinting pattern to reveal genetic diversity and the abundance of microbial communities [13,14,15,16,17,18,19,20]. Quantitative PCR (qPCR) is a technology used to monitor the number of copies of a targeted DNA molecule [21]. Redundancy analysis (RDA) is a multivariate regression analysis used to explain the variation in a set of response variables by another set of explanatory variables. The method has been widely applied to extract and summarize the relationships between microbial community composition and environmental factors [22]. Regression analysis is a statistical method which is utilized for studying the dependence relationships between dependent variables and independent variables [23].
Few studies have aimed to explore the influences of physico-chemical parameters on Actinomycetes community composition changes during composting. Therefore, we measured eight physico-chemical parameters, investigated the Actinomycetes community composition by denaturing gradient gel electrophoresis (DGGE), analysed the relationships between the parameters and the Actinomycetes communities by RDA, determined the quantity of Actinomycetes 16S rDNA gene by qPCR and studied the correlations between these parameters and Actinomycetes 16S rDNA gene abundance with regression analysis. It was hypothesized that quantifying and prioritizing the relationships between physico-chemical parameters and Actinomycetes community composition would provide a better understanding of agricultural waste composting for the scientific community and provide some useful references for relevant research and practice.
2. Material and Methods
2.1. Composting and Sampling
Soil collected from the surface 10–15 cm of soil on Yuelu Mountain (Changsha, Hunan, China) was sieved through a 40-mesh screen to remove coarse plant debris and was added to a composting system to increase microbial species and provide nutrients for microorganisms. The vegetables used in this study included Chinese cabbage, lettuce and celery, which were chopped into 1–2 cm pieces and used as easily degradable organic material for microorganism. Rice straw, as the typical agricultural waste, was gathered from the suburb of Changsha, Hunan, China, which was used as a difficult-degradable organic material, after air-drying it was cut into 1–2 cm lengths. As a low total carbon/total nitrogen (C/N) raw material, bran was used to adjust the initial C/N ratio of composting which was the optimal C/N ratio for the microbial activity [22]. The characteristics of the above materials are shown in Table 1.
The pilot-scale composting pile was placed in a special experimental container (50 L) with good heat preservation. All materials for composting weighed about 10 kg (dry weight) in total. Materials were mixed adequately at a calculated ratio and packed loosely according to a method reported by Zeng [24]. Rice straw, vegetables, bran and soil were mixed at a ratio of 11:3:2:8 (dry weight) to obtain a mixture with a C/N ratio of about 30:1 and the moisture content was adjusted to about 60%, which was considered to be the optimal range for microorganism activity [2]. The experimental period was 45 days. About 200 g composting samples were collected on days 1, 2, 3, 4, 5, 6, 9, 12, 15, 18, 21, 27, 33, 39 and 45, respectively. In order to minimize sampling errors, three subsamples were collected from different sections of the pile (the top, middle and bottom), respectively. The three subsamples were mixed well and then divided into two parts: one for physico-chemical parameter analysis and the remaining were stored immediately at −20 °C for total DNA extraction. After sampling, the moisture content of the composting pile was adjusted using the method described by Zhang [25]. In order to supply enough oxygen, the materials were turned adequately every day during the first 12 days and once a week after that until the end.
2.2. Physico-Chemical Parameters Analyses
To ensure the accuracy of the results, the composting pile was performed in triplicate and the results were presented as the mean of these data. The C/N ratio, ambient temperature, moisture content, pH, WSC and pile temperature were measured according to our previous study [22]. Besides that, the ammonium and nitrate of materials were extracted with a 2M KCl solution (materials:solution, 1:5) and its concentrations were determined using the Continuous Flow Analyzer (A33, SEAL Analytical, Norderstedt, Germany). For all measurements, three parallels were tested.
2.3. DNA Extraction and PCR-DGGE
The protease K and cetyltrimethyl ammonium bromide (CTAB) method was used to extract the total DNA of all samples [26]. The crude DNA was purified with a TIAN quick Midi Purification Kit (TianGen, Beijing, China) following the manufacturer’s recommendations. A fragment length of about 433 bp of Actinomycetes was amplified by Nested PCR. In the first step, Actinomycetes-specific primer F243 and R1378 were used to amplify Actinomycetes 16S rRNA genes from positions 226 to 1401. Then 2 µL amplification products in the first step were used as the DNA template. Primer F984GC and R1378 were used to amplify the fragment from positions 968 to 1401 with a length of about 433 bp that is appropriate for DGGE. The reaction conditions of the two amplification steps were followed according to the protocol described by Das [27]. The details of the primers used in Nested PCR are shown in Table 2.
The 40 µL amplification products from the second PCR process of each sample were used for DGGE analysis. The analysis was performed with a DCodeTM Universal Detection System apparatus. The amplification products were loaded into 8% (w/v) polyacrylamide gels in 1 × TAE with a denaturing range from 30 to 80%. The gel electrophorese was run at 60 °C and 80 V for 12 h, then the gel was stained for 30 min by using SYBR Green I nucleic acid gel stain (Molecular Probes, Carlsbad, CA, USA) for 30 min, photographed by a UV Transilluminator and analysed using Quantity One software (version 4.5, Bio-Rad, Hercules, CA, USA).
2.4. qPCR Analysis
16S rRNA gene copy numbers of Actinomycetes were measured with real-time qPCR with the primers F984 and R1378 in iCycler IQ5 Thermocycler (Bio-Rad). The qPCR reaction mix contained 0.5 μL of each primer (10 μM), 10 μL of 2 × SYBR real-time PCR pre-mixture (Invitrogen, Carlsbad, CA, USA), 0.5 μL of BSA (10 mg mL−1), 1 μL of 10-fold dilution of DNA extract and was adjusted to 20 μL with sterile deionized water. The amplification was performed by using the following conditions: 94 °C for 3 min, followed by 40 cycles of 20 s at 94 °C, 30 s at 55 °C and 30 s at 72 °C. Data was obtained at 72 °C. A known copy number of plasmid of the Actinomycetes 16S rDNA gene was ten-fold serially diluted and used as a template in qPCR for the standard curve generation. For all the qPCR experiments, duplicate runs were used in all samples.
2.5. Data Analysis
The One-Way Analysis of Variance (ANOVA) method was used to analyse all physico-chemical parameters and to test the significant differences between the data. Data were obtained as average values of the triplicates, the maximum difference between which was 5%.
After subtracting a rolling disk background value for polyacrylamide gels and each lane, a DGGE banding profile for Actinomycetes communities was digitized and calculated the relative contribution of each band to the total band intensity for each lane by the Quantity One software [28]. Multivariate relationships between Actinomycetes community composition and eight physico-chemical parameters were determined by Canoco (version 4.5, Centre for Biometry, Wageningen, The Netherlands).
Different magnitudes of the units existed in all the physico-chemical parameters, so the multivariate analyses could not be carried out directly. SPSS Statistics (version 11.5, SPSS, Chicago, IL, USA) software was used to standardize those parameters [2,29]. A detrended correspondence analysis (DCA) was used first to analyse the data of Actinomycetes communities to decide whether to use either a linear or unimodal response model. The analysis showed that the first DCA ordination axis length was 2.448. Therefore, the redundancy analysis (RDA) method, based on a linear response model, was used to analyse the response of the Actinomycetes community composition to the eight physico-chemical parameters [30].
The RDA was carried out with default settings [2]. Regression analysis was performed to determine the correlations between each of the eight physico-chemical parameters and the quantity of Actinomycetes 16S rDNA gene abundance by SPSS (version 11.5).
3. Results and Discussion
3.1. The Changes of Physico-Chemical Parameters
Figure 1 described the physico-chemical parameters changes over 45 days. As shown in Figure 1a, the C/N ratio decreased gradually throughout the whole process from 31.15 to 16.62 due to the fact that the organic matter was degraded gradually by the microorganism. Values of C/N ratio lower than 20 at day 27 were considered the threshold value for composting maturity [22]. During the first 6 days, the ammonification and mineralization process produced plenty of ammonium, so the pH increased from 6.74 to 8.87. Then the pH decreased slowly to 8.13 until day 21 as a result of the release of H+ derived from the nitrification process and the volatilization of ammoniacal nitrogen. Due to the abundant growth and metabolism of the microorganism, easy-metabolizable materials were degraded quickly and the pile temperature increased rapidly to 56 °C during the first 4 days and was maintained above 50 °C for 12 days, which is long enough to suppress a pathogenic microorganism, then decreased gradually to become stable at a little higher than ambient level (Figure 1b). Moisture content was kept at about 60% during the first fermentation phase and at about 45–50% during the second fermentation phase, which was considered the optimal range for microbial activity [25].
WSC increased from 217.5 mg kg−1 to 245.6 mg kg−1 during the first 3 days and then decreased gradually to 57.6 mg kg−1 at the end (Figure 1c). The ammonium and nitrate concentration showed an almost opposite variation trend. Nitrate concentration increased rapidly to 1166.8 mg kg−1 (dry weight) on the 2nd day and ammonium increased significantly to a maximum of 2122.2 mg kg−1 (dry weight) on the 5th day respectively, then the nitrate decreased quickly to a minimum of 398.34 mg kg−1 and increased gradually to 1353.4 mg kg−1, after the 5th day ammonium decreased gradually to 296.8 mg kg−1 (Figure 1d). The pH, pile temperature and ammonium all increased during the early stage of composting, which was explained by the abundant growth of the microorganisms which need a large amount of nutrients to maintain their metabolism. Therefore, easy-metabolizable materials were degraded and released heat. Meanwhile, after microorganism degradation, organic nitrogen was transformed to ammonium by the ammonification and to nitrate by the mineralization during the early stage of composting.
3.2. DGGE Analysis and Quantitative Analysis
Composting is a process controlled by microorganisms. PCR-DGGE is a powerful technology to monitor the composition changes in a microbial community that is affected by environmental factors [31]. The DGGE gel result of Actinomycetes community composition is shown in Figure 2. The changes of the band intensity at the same horizontal position suggested a relative change of the species abundance in different samples [32]. Each band of the DGGE gel image represents a group of Actinomycetes species. Bands of Actinomycetes 16S rDNA sequences in the same horizontal position mean they have a similar melting characteristic [33]. Figure 2 indicates that the Actinomycetes species were dynamic and the number of Actinomycetes species during the last phase of composting was more than in other phases.
The Actinomycetes 16S rDNA copy numbers measured by qPCR are shown in Figure 3. The numbers increased quickly at the beginning of the composting, decreased rapidly between day 12 and 15 and then increased slowly until the composting process was ended.
3.3. Redundancy Analyses and Regression Analysis
Redundancy analysis was used to analyse and digitalize the extent to which the eight physico-chemical parameters affected the Actinomycetes community’s composition. Therefore, the influences between physico-chemical parameters and the Actinomycetes community’s composition during composting was analysed by the relative band intensities of the community profile data. The results show that these parameters all influence the composition of the Actinomycetes community because the significance of the first axis (p = 0.006) and all canonical axes (p = 0.004) were significant by Monte Carlo permutation tests carried out with default settings. The first two canonical axes explained 66.4% and 7.8% of the variation in Actinomycetes community composition, respectively. All four significant canonical axes and all canonical axes explained 84.2% and 86.6% of the changes in the Actinomycetes community composition, respectively (Table 3).
Regression analysis was used to analyse the relationships between these physico-chemical parameters and Actinomycetes 16S rDNA gene abundance by SPSS (version 11.5). The quantitative curve of Actinomycetes 16S rDNA gene abundance showed a negative linear relationship (R2 = 0.407, p < 0.01) with the moisture content and other parameters had no obvious correlation relationship with the variation tendency of quantitative curve, which indicated that moisture content had a more important influence on Actinomycetes communities than other parameters during composting (Figure 4).
In this study, the eight physico-chemical parameters accounted for 86.6% of the variation in the composition of the Actinomycetes communities by RDA. The variation of about 13.4% that remained unexplained may be related to other physico-chemical parameters, such as the content of humus and the interaction between microorganisms, which were not measured in this study. Only moisture content showed a significant correlation with the Actinomycetes 16S rDNA gene abundance by regression analysis. We suggest that future studies of Actinomycetes should link Actinomycetes function (e.g., enzyme production) with physico-chemical parameters to explore the interaction effect between them.
3.4. Forward Selections and Variation Partitioning
In this study, in order to make clear the influences of physico-chemical parameters on composition changes in the Actinomycetes community, we sorted the significant parameters based on their relative influence by variation partitioning analysis and expounded changes during composting. Forward selection was carried out to screen the significant impact factors (p < 0.05), which had a statistically significant influence on Actinomycetes community composition. The results indicated that the variation in Actinomycetes community composition was influenced significantly by WSC, moisture and pH (p < 0.05). In order to quantify the effects of these three significant factors on the variation of the actinomycetes community composition, variation partitioning analysis was carried out to conduct further analysis [34]. The results are shown in Table 4; 45.4% of the variation (p = 0.018) was explained statistically by the RDA model in all of the Actinomycetes community composition during composting. WSC, moisture content and pH explained 2.8% (p = 0.002), 19.6% (p = 0.002) and 5.3% (p = 0.03) of the variation of Actinomycetes community composition, respectively and the three significant impact factors explained 17.7% of variation in all, without the effect of other parameters. The positions of each sample with the first two environmental axes are shown in Figure 5, providing a more intuitive expression of the composting condition of each sampling time.
Partial RDAs based on the Monte Carlo permutation (n = 499) kept only the significant parameters in the models. For each partial model, the other significant parameters were used as covariates. F and p values were estimated using Monte Carlo permutations. The sum of all Eigen values for both partial RDAs were 1.000.
WSC is the predominant factor influencing bacterial and fungal community composition [2]. Straathof [35] and Maeda [36] drew a common conclusion that dissolved organic carbon concentration is the main factor affecting microbial communities in the composting process. In this research, WSC showed a significant correlation (p = 0.002) with Actinomycetes community composition. Thus, it can be seen that WSC has a great influence on microbial community composition. It was reported that pH was a remarkably influential factor which had different effects on microbial communities under different conditions during the composting process [11]. In this study, the pH (p = 0.03) explained 5.3% of the variation of Actinomycetes community composition. In addition, moisture content induced a significant modification in the Actinomycetes community composition (19.6% of the variation, p = 0.002), which suggested that Actinomycetes communities were more sensitive to moisture content fluctuation than other parameters. Ryckeboer [6] also reported that moisture content was an important factor for Actinomycetes. However, it was reported that WSC and pile temperature were the main factors affecting bacterial and fungal community composition, respectively [2]. That is to say, compared with bacterial and fungal communities, the composition of the Actinomycete community was more sensitive to moisture content.
Ammonium and nitrate have a large influence on bacterial communities [2]. However, no significant relationships with Actinomycetes communities were found in this study. Although several researchers [22,37,38] have reported that pile temperature had a significant impact on microbial activity and biomass during composting, no significant correlation between pile temperature and Actinomycetes communities were found during composting in this study. Maybe the influence of pile temperature on Actinomycetes communities were covered by moisture content because of adding deionized water to compost at regular intervals. Ambient temperature had almost no influence on Actinomycetes communities because of less direct interaction.
4. Conclusions
The pile entered the maturity stage at 27 days. These eight physico-chemical parameters statistically explained 86.6% variation of the Actinomycetes community composition during agriculture composting. Moisture content, pH and WSC had the predominant effect, statistically explaining 19.6%, 5.3% and 2.8% variation, respectively and 45.4% variation in all of the Actinomycetes community composition during composting. Only moisture content showed a negative linear relationship with the quantity of Actinomycetes 16S rDNA gene abundances. Actinomycetes species were dynamic and the number of Actinomycetes species during the last phase of composting was more than other phases. In addition, compared with bacterial and fungal communities, the composition of the Actinomycete community was more sensitive to moisture content.
Author Contributions
Y.L. (Yuanping Li), Y.C. (Yanrong Chen) and Y.C. (Yaoning Chen) conceived and designed the experiments; Y.L. (Yuanping Li) and Y.C. (Yanrong Chen) performed the experiments; Y.L. (Yihuan Liu), Z.P., Z.Z., R.X. and S.W. analyzed the date; Y.C. (Yaoning Chen), Y.W. and C.Z. contributed materials and analysis tools; Y.L. (Yuanping Li) and Y.C. (Yanrong Chen) wrote the paper; Y.W. and C.Z. provided comments to the paper.
Funding
This study was financially supported by the Research Foundation of Education Department of Hunan Province (17C0305, 17K019), Natural Science Foundation of Hunan Province, China (2017JJ2020), Outstanding Youth Project of Hunan Province, China (16B049) and the Natural Science Foundation of Hunan Province (13JJB002).
Conflicts of Interest
The authors declare no conflicts of interest.
Figures and Tables
Enlarge this image.Figure 1. Changes of physico-chemical parameters during composting process: (a) pH and C/N ratio; (b) ambient and pile temperature; (c) WSC and moisture content; (d) ammonium and nitrate.
Enlarge this image.Figure 1. Changes of physico-chemical parameters during composting process: (a) pH and C/N ratio; (b) ambient and pile temperature; (c) WSC and moisture content; (d) ammonium and nitrate.
Enlarge this image.Figure 2. DGGE profiles of amplified Actinomyces 16S rDNA fragments from the compost samples. The numbers refer to the sampling days.
Enlarge this image.Figure 3. Actinomyces 16 S rDNA gene abundance in different sampling time during agricultural waste composting.
Enlarge this image.Figure 4. Negative linear relationship between Actinomyces 16S rDNA abundance and moisture content.
Enlarge this image.Figure 5. DGGE band data redundancy analysis for Actinomyces species. Solid lines with filled arrows and grey dotted lines with filled arrows represent significant composting parameters and supplementary parameters, respectively. Samples are represented by open circles and sample numbers refer to the sampling days.
Basic characteristics of the composting raw materials.
| Materials | Moisture Content % | Organic Materials % | C% | N% | C/N |
|---|---|---|---|---|---|
| Rice straw | 0.1173 | 0.7380 | 0.4280 | 0.0088 | 48.7600 |
| Bran | 0.1406 | 0.8174 | 0.4741 | 0.0412 | 11.5000 |
| Vegetables | 0.7906 | 0.1676 | 0.0972 | 0.0050 | 19.5500 |
| soil | 0.2613 | 0.0950 | 0.0551 | 0.0024 | 22.6700 |
All the above data are dry sample data.
Table 2Details of the primers.
| Primer | Primer Sequence |
|---|---|
| F243 | 5′-GGA TGA GCC CGC GGC CTA-3′ |
| R1378 | 5′-CGG TGT GTA CAA GGC CCG GGA ACG-3′ |
| F984GC | 5′-GC-AAC GCG AAG AAC CTT AC-3′ |
Redundancy analysis results of Actinomyces DGGE profile.
| Axis | Eigen Value | Species-Environment Correlation | Cumulative % Variation of Species | Cumulative % Variation of Species-Environment | Sum of All Canonical Eigen Values |
|---|---|---|---|---|---|
| Axis 1 | 0.664 | 0.986 | 66.4 | 76.7 | 0.866 |
| Axis 2 | 0.078 | 0.88 | 74.2 | 85.7 | |
| Axis 3 | 0.072 | 0.912 | 81.4 | 94 | |
| Axis 4 | 0.028 | 0.883 | 84.2 | 97.3 |
Monte Carlo significance tests: sum of all Eigen values, 1.000; significance of first canonical axis, F-ratio = 11.869, p = 0.006; significance of all canonical axes, F-ratio = 4.841, p = 0.004.
Table 4The influences of the significant parameters on Actinomyces community composition were tested by Eigen values, F values and p values which were obtained from the partial RDA.
| Parameters Included in the Model | Eigen Value | % Variation Explains Solely | F Value | p Value |
|---|---|---|---|---|
| moisture | 0.196 | 19.6 | 7.79 | 0.002 |
| pH | 0.053 | 5.3 | 2.71 | 0.03 |
| WSC | 0.028 | 2.8 | 6.58 | 0.002 |
| All the above together | 0.454 | 45.4 | 3.047 | 0.018 |
© 2019 by the authors.
Details
; Wu, Yanxin 2 ; Zhang, Chun 1 ; Peng, Zhen 2 ; Liu, Yihuan 2 ; Wang, Sha 2 ; Xu, Ran 2 ; Zeng, Ziping 2 1 College of Municipal and Mapping Engineering, Hunan City University, Yiyang 413000, China
2 College of Environmental Science and Engineering, Hunan University, Changsha 410082, China; Key Laboratory of Environmental Biology and Pollution Control (Hunan University), Ministry of Education, Changsha 410082, China