3.1. Comparison of temperature, moisture and pH of samples
In Fig. 1, the pile temperature of wheat straw + swine tissues treatment (ST) rapidly increased and reached the thermophilic phase in the first 12 days of the process. The thermophilic phase was shorter in woodchips + swine carcass (WS) and rice husk + swine carcass (RH) treatments than in wheat straw + swine carcasses(ST) treatment. Moreover, the highest temperature in three composting treatments was in the order of ST (55°C) > WS (42.6°C) > RH (30.8°C). This might be due to the initial C/N ratio of ST treatment (close to 30) (Table S1), which was determined as a better condition for reproduction of microorganisms (Hashemi and Han, 2019).
In Fig.2a, the sharpest decrease of moisture content among the three treatments was observed in the RH treatment in the first 30 d of composting, probably due to the compensation for the evaporation of moisture via the production of metabolic heat and water by biodegradation. After 30 d, the moisture continued to decline in the ST treatment, probably due to its longer thermophilic stage, resulting in the greatest decreasing rate of moisture content among the three treatments. In Fig.2b, the pH value was shown to decrease rapidly in the RH treatment in the first 12 d, probably due to the microbial activity on easily biodegradable organic matter and the release of organic acids in the materials (Robledo-Mahon et al., 2019; Tang et al., 2020; Lopez-Gonzalez et al., 2015). As composting progressed, the pH increased to around 7.0 at the end of the process, indicative of proper stabilisation of organic matter (Robledo-Mahon et al., 2019; Liu et al., 2018).
3.2. Comparison of total organic carbon of samples
The total organic carbon (TOC) was detected to decrease gradually during the composting process in all the three treatments. The initial TOCs of 48.84%, 41.75% and 38.69% decreased to 44.32%, 36.32% and 32.42% in the final composts of WS, RH and ST, respectively (Fig. 3). The TOC decrease percentage was in the order of ST (15.9%) > RH (12.99%) > WS (9.32%). Compared with WS and RH, the ST feedstock exhibited slightly faster decomposition based on TOC reduction, probably attributed to more easily degradable organics contained in ST than in WS and RH (Qian et al., 2014).
3.3. Comparison of cellulose and lignin content of samples
Figure 4 shows the cellulose and lignin contents of the three treatments throughout the 78-day natural composting period. In Fig. 4a, the final cellulose degradation rate was seen to be in the order of ST (16.25%) > WS (14.08%) > RH (13.01%). In Fig. 4b, the final lignin degradation rate was shown to be ordered as ST (45.05%) > RH (25.33%) > WS (0.79%). Both the lignin and cellulose degradation rates were significantly higher in ST than in WS or RH. A more favorable condition in ST may improve the microbial activity of cellulose-decomposing microorganisms (Sun et al., 2016; Yu et al., 2007; Zhao et al., 2016). The aforementioned results indicate that wheat straw could significantly promote lignocellulose degradation during swine tissue composting.
3.4. Comparison of nitrogen profile of samples
With the degradation of swine tissues, the total nitrogen content of the three groups all showed an upward trend (Fig. 5a). At the end of composting, the total nitrogen content of the straw group was the highest. It may be that straws have high nitrogen content and are rich in phosphorus, potassium, calcium, and magnesium, which are suitable for the growth of microorganisms and promote the degradation of organic matter
In Fig. 5b, the ammonium nitrogen (AN) content of ST and RH treatments showed a increase and reach a peak value, followed by a decrease throughout the composting process. Previous studies have indicated that the increase of ammonium nitrogen (AN) can be attributed to the mineralization and ammonification of organic nitrogen. up to 53d, the straw group had the highest ammonium nitrogen content among 3 groups, which further confirmed that the conversion efficiency of organic nitrogen was high and the composting process was good at the straw group. The ammonium nitrogen content of the woodchips group gradually increased, indicating that the composting process of the woodchips group was slow, but the woodchips has a larger void ratio, which can absorb NH4+ (Hashemi et al., 2016), resulting in the highest final ammonium nitrogen content.
3.5. Comparison of bacterial communities
As shown in Table 1, the sequences of each sample were clustered as Operational Taxonomic Units (OTU) with over 97% identity. Among the three treatments, ST showed the highest OTU number in the samples at thermophilic stage, while the lowest OTU number in initial stage. In terms of Shannon index, ST showed a greater increase and decrease trend than WS or RH. The decrease of Shannon index from ST1 to ST2 might be due to the death of some species during the mesophilic phase, and the marked rise from ST2 (5.78) to ST3 (8.82) was related to the contribution of sufficient nutrients to the growth of some thermophilic or thermotolerant microorganisms (Sun et al., 2019) (Meng et al., 2019). The obvious drop of Shannon index after thermophilic phase was in line with the changes of Chao 1 index. In general, the highest diversity was shown by ST3 in the ST treatment, RH3 in the RH treatment, and WS4 in the WS treatment.
Table 1
Effects of different treatments on Observed Sequences, OTU number, estimated indices (Simpson, Chao 1, ACE and Shannon) and coverage in five sampling stages.
Sample
|
Sequence
|
OTUs
|
Simpson
|
Chao1
|
ACE
|
Shannon
|
WS1
|
45760
|
155
|
0.976902
|
1004.21
|
1046.8
|
7.47
|
WS2
|
44242
|
199
|
0.975829
|
1184.18
|
1236.44
|
7.49
|
WS3
|
31531
|
178
|
0.990941
|
2024.78
|
1970.42
|
8.84
|
WS4
|
34659
|
262
|
0.99608
|
2591.25
|
2602.74
|
9.52
|
WS5
|
34223
|
218
|
0.991009
|
2181.46
|
2333.1
|
9.15
|
RH1
|
38100
|
153
|
0.970149
|
967.56
|
1018.12
|
7.01
|
RH2
|
30279
|
193
|
0.979968
|
1302.64
|
1397.27
|
7.62
|
RH3
|
31768
|
247
|
0.992569
|
1953.25
|
2061.54
|
9.15
|
RH4
|
28314
|
206
|
0.968353
|
1325.03
|
1327.73
|
8.02
|
RH5
|
36626
|
234
|
0.991915
|
1942.27
|
1911.68
|
8.78
|
ST1
|
30948
|
169
|
0.968464
|
1125.01
|
1211.52
|
7
|
ST2
|
34252
|
173
|
0.93199
|
711.65
|
741.55
|
5.78
|
ST3
|
50685
|
399
|
0.990531
|
1979.12
|
2081.65
|
8.82
|
ST4
|
34685
|
266
|
0.989882
|
1712.15
|
1682.66
|
8.46
|
ST5
|
32329
|
194
|
0.991996
|
1345.6
|
1360.71
|
8.67
|
Note: Chao1 richness estimation index (Chao1) indicates the number of species in the community; the ACE richness estimation index (ACE), the number of species in the community; Shannon diversity index (Shannon), the richness and evenness of the community; Simpson diversity index (Simpson), community diversity. The first column in the table is the sample name, followed by indexes of Observed Sequences, OTU number, Chao1, ACE, Shannon, and Simpson coupled with the results of samples in the same sequencing depth. 1, 2, 3, 4, 5 represent samples at five different periods, including the initial phase, the warming phase, the high temperature phase, the cooling phase, and the maturity phase. WS = woodchips + swine carcasses; RH = rice husk + swine carcasses; ST = wheat straw + swine carcasses. |
3.6. Comparison of bacterial community composition at phylum and genus level
Figure 6 presents the classification of 16S rRNA gene sequences at the phylum and genus level. For all the samples at the phylum level (Fig. 6a), the most dominant phyla included Proteobacteria, Actinobacteria, Bacteroidetes, Firmicutes, Cyanobacteria, Chloroflexi, accounting for over 99% of the total 16S rRNA gene sequences in each sample. The first four dominant taxonomic phyla were also found in composting of other wastes (Antunes et al., 2016; de Gannes et al., 2013; Wei et al., 2018).
Proteobacteria was incredibly diverse and contained members of great importance to carbon, sulfur, and nitrogen cycles of the planet (Wei et al., 2018; Zhong et al., 2018; Yang et al., 2019). The abundance of Proteobacteria showed a downward trend in WS and RH treatments, but a trend of first rise and then fall in the ST treatment at the thermophilic phase, possibly due to the great importance of Proteobacteria phylum to global C, N and S cycling (Burges et al., 2020).
Actinobacteria plays a important role in degradation of refractory cellulose and lignin (Partanen et al., 2010; Peters et al., 2000; Su et al., 2015). Bacteroidetes plays a major role in organic matter degradation and C cycling (Wang et al., 2018); Bacteroidetes could break down lignocellulose into short chain fatty acids (Dodd et al., 2011; Zhong et al., 2018). Firmicutes is thought to play a major role in lignocellulose degradation (Pankratov et al., 2011). The relative abundance of Bacteroidetes and Firmicutes was both higher in ST than in WS or RH, implying more rapid degradation of cellulose and lignin in ST (Fig. 4a-b). Cyanobacteria is the main N2 fixing organism in freshwater ecosystems (Li et al., 2019). Cyanobacteria showed higher abundance in ST than in WS or RH, providing support for N pollution control, remediation, and management. Chloroflexi contains aerobic and anaerobic thermophiles, filamentous anoxygenic phototrophs, and anaerobic organohalide respirers (Maymo-Gatell et al., 1997; Xu et al., 2019). Chloroflexi exhibited the highest relative abundance in ST5 among all the samples, promoting rapid biodegradation of organic substances and then humification. Planctomycetes is considered as slow-growing decomposers of organic matter and has a unique anaerobic ammonium oxidation trait (Kulichevskaya et al., 2012; Zhong et al., 2018). Planctomycetes had the highest abundance in WS5 among all the samples, and this phylum may contribute to net nitrification in the compost piles.
Figure 6b shows the top 20 genera in the composting samples. Among them, Pseudomonas, Brevundimonas, Acinetobacter, Devosia and Rhizobium belong to Proteobacteria. Streptomyces, Brevibacterium, Saccharoporyspora, Microbacterium, Sanguibacter and Brachybacteriumd belong to Actinobacteria. Sphingobacterium belongs to Bacteroidetes. Pseudomonas is widely distributed in nature and can decompose complex polymers such as lignocellulose (de Gannes et al., 2013). The relative abundance of Pseudomonas was higher in ST than WS or RH, probably due to its longer thermophilic phase and higher degradation rate of lignocellulose than either of them. Streptomyces and Acinetobacter began to emerge at the initial phase of WS treatment and reached the highest level at maturity phase. Streptomyces was a dominant bacterium during composting of swine carcasses and woodchips (Yang et al., 2019), which was related to the consumption and assimilation of ammonia (Kim et al., 2013; Yang et al., 2019). A comparison of these three treatments revealed the highest relative abundance of Brevibacterium in RH3 sample, which was consistent with the rise of total nitrogen content at the high temperature perid. Brevibacterium was related with total nitrogen and ammonium nitrogen content in swine carcass composting (Yang et al., 2019). Overall, the total abundance of the top 20 dominant genera was the higher in ST than in WS or RH, indicating that the wheat straw treatment environment is more suitable for microbial growth. The dominant bacteria were Pseudomonas, Staphylococcus and Sphingobacterium in ST treatment, Streptomyces Acinetobacter and Sphingopyxis in WS treatment, and Sanguibacter, Pedobacter and Gordonia in RH treatment.
3.7. Correlation analysis between bacterial communities and selected factors
Environmental factors are the main drivers for the development of microbial communities (Maeda et al., 2010). The potential effects of various physicochemical parameters (pH, AN, MC, temperature, TC, TN, cellulose, hemicellulose and lignin) on the bacterial community structure throughout the composting process were evaluated by Spearman analysis of the relationship between environmental factors and top 20 genera (Fig. 7). The variables of AN, MC, TN, hemicellulose and lignin and composting temperature were shown to have different (positive/negative) impacts on bacterial communities. Pseudomonas was positively related to MC and hemicellulose, but negatively related to temperature and AN. Brevibacterium was positively related to temperature, TN and AN. Saccharoporyspora showed a negative correlation with temperature. Brevundimonas showed a significant and positive correlation with pH and cellulose. Devosia showed a positive correlation with AN and temperature. Those observations corroborated our analysis results of physicochemical parameters. In terms of compost product, Pseudomonas, Brevibacterium, Devosia, Staphylococcus, Microbacterium, Sanguibacter, Brachybavterium and Promicromonospora all showed a positive correlation with N content, suggesting the high quality of the compost product. Among them, Brevibacterium, Devosia, Sanguibacter, Brachybavterium and Promicromonospora all showed a positive correlation with temperature, suggesting the high maturity of the compost product.