Changes in physicochemical parameters during composting
During the mesophilic phase, temperature of piles rose quickly and steadily, and EG was slightly higher than the CK (Fig. 1a). After turning over piles (27 d), the temperature level increased rapidly again and subsequently entered into the thermophilic stage (>50℃). Though the temperature trends of two treatment groups were similar, temperature in the EG was significantly higher compared with the CK. EG reached the highest temperatures on day 35 at 67.3 °C and kept above 50 ◦C for 27 days, by contrast, CK reached the highest temperature on day 44 at 53.4 ◦C and kept above 50 ◦C for 8 days. Subsequently, the temperature of two treatment groups dropped on days 43 and 45, respectively, and entered the cooling stage. Results indicated that EG probably possessed more available functional microorganism led to more organic waste was metabolized and released more energy. Appropriate high temperature and keep a longer thermophilic phase is killing not more pathogens and weed seeds, but also making compost product safer.
The pH values of two treatment groups showed both a decrease first and then increase as compost time proceed (Fig.1b). The main point for pH decreased may be due to the microorganisms rapidly decomposed large quantity of easily degradable organic matter that caused organic acids produced (Gajalakshmi et al. 2008) and consumed simultaneously a part of nitrate-nitrogen (Guardia et al. 2009). The pH values of two treatment groups decreased after entering the thermophilic phase, and which was significantly lower in the EG compared to the CK. It was no surprise that on the one hand high temperatures generally cause the NH4+-N to volatilize to NH3, on the other hand, the thermophilic microorganisms worked industriously to decompose organic matter and produce small molecule acids. During the cooling-off period, undegraded organic matter, such as proteins, was generally further decomposed and produced NH4+-N. Consequently, the pH values of two treatment groups the period increased slightly, and which was similar to previous studies (Meng et al. 2018a, b).
The EC values of two treatment groups were all more than 4.0 mS/cm in the whole composting process (Fig. S1), and seemed to be related with the local characteristic of the soil environment (Bi et al. 2020). The problem should give rise to people's attention because a high soluble salt content will inhibit the growth of crops. Moreover, the content of available nitrogen in the EG was higher than the CK since microbial inoculation might reduce the volatilization of ammonia and increase the available nitrogen content (Fig. S1).
Evolution of lingo-cellulosic fractions during composting
The activity of lignocellulosic decomposition depends on the species and abundance of lignocellulose-degrading microorganisms present in the mixture. The cellulose, hemicellulose, and lignin contents of two treatment groups gradually decreased with the composting proceed (Fig. 1d-f). It was obvious that the EG produced the significantly higher (P < 0.05) cellulose, hemicellulose, and lignin degradation rates. Specifically, the cellulose, hemicellulose, and lignin degradation rate of the EG increased from 53.3% to 70.0%, 50.2% to 61.3%, and 46.4% and 58.9% respectively in nearly finished composting (60 d) compared to the CK.
The degradation of cellulose and lignin occurred a lot 35.2% to 32.9.0% and 25.1% and 33.9%, respectively) in two treatment groups during the thermophilic phase, which was consistent with previous reports (Mei et al. 2020; Zhu et al. 2021) and Xiao et al. (2009) also found that cellulose degradation was stronger during high temperature period. EG generated the higher temperature and kept a longer thermophilic period, which was helpful for improving the decomposition efficiency of cellulose and lignin. In addition, cellulose was tightly cross-linked with lignin in the lignocellulose matrix, so the internal lignin components will be more exposed when cellulose was significantly removed. Therefore, it was more effectively promoted the lignocellulolytic enzymes secreted by microorganisms to play their role in the EG.
Seed germination index (GI) analysis
Seed germination index (GI) was used for quickly and efficiently assessing the compost maturity and has been widely accepted by people (Hussain et al. 2018). Composts with a GI value > 80.0% were considered to be mature according to Bernal et al. (2009). GI was analyzed with aqueous extracts of fresh samples using Chinese cabbage. The GI of the EG and the CK enhanced gradually along with composting process and eventually increased from 33.6% and 35.1% at day 0 to 90.7% and 83.0% at day 60, respectively (Fig. 1c). Results showed that addition of microbial inoculation had a significant impact on GI of the final compost product, the final products of EG was mature and non-phytotoxic, which in line with the national normative requirement (Luo et al. 2017).
Changes of bacterial and fungal community diversity during composting
After quality filter and potential chimeras were removed, a total of 3,799,998 high-quality bacterial and 2,111,338 fungal sequences were generated respectively for 48 samples across two treatment groups. Regarding the β-diversity, Principal Component Analysis (PCA) revealed a significant effect of the microbial inoculation on bacterial and fungal community, the difference became more obvious with compost time (Fig. 2a and c). The α-diversity was obtained for Shannon's diversity indexes, Chao1, and observed OTUs. The Shannon's diversity indexes were in general different between compost samples in the EG and the CK (Fig. 2b and d). Compared with bacteria, the Shannon's diversity index of fungi has risen notably on day 48. For bacteria and fungi, it could observe that EG samples showed significantly higher values for all the diversity indices as compared to the CK samples (Fig. S2). Moreover, although the α-diversity of the EG was relatively more stable from initial to mesophilic period (27 d), which changed markedly in both the EG and the CK in the thermophilic period. Venn diagram exhibited that EG and CK samples possessed 1214 (8.1%) and 933 (6.0%) bacterial OTUs shared by four different compost stages, respectively (Fig. S3). Meanwhile, they possessed 64 (4.4%) and 52 (3.5%) fungal OTUs shared by four compost stages, respectively. Results suggested that EG could slightly enhance the core bacterial OTUs, but not the core fungal OTUs.
At the phyla level, the abundances of the top nine phyla represented 95~97% of the total bacterial communities, including Proteobacteria (21.8-52.4%), Bacteroidetes (14.5-39.0%), Firmicutes (4.0-37.3%), and Actinobacteria (4.2-9.6%) and the remainder belonged to the phyla Spirochaetae, Chloroflexi, Gemmatimonadetes, Planctomycetes and Fibrobacteres (Fig. 3a). Because of the copiotrophic strategies of the Proteobacteria and Bacteroidetes, they usually showed the rapid growth response to resource availability (Fierer et al. 2007). Thus, Proteobacteria and Bacteroidetes increased first and then decreased during composting. By contrast, Firmicutes increased markedly in the thermophilic phase, which was able to secrete various extracellular thermostable enzymes and degrade some macromolecular substrates (protein, pectin, and cellulose, etc.). The addition of microbial inoculation had not a significant impact on the abundances of Actinobacteria, but Bacteroidetes on day 27. On day 48, Firmicutes and Gemmatimonadetes were more abundant in the EG than the CK. The previous findings indicated that they are important bacteria for anaerobic fermentation, decomposing organic matter to hydrogen or acetic acid (Cardinali-Rezende et al. 2012), which also matched well with the levels of cellulose and lignin degradation in the compost process.
Ascomycota was the most well-represented fungal division during the composting process, composing more than 87% of the fungal species in the thermophilic stage, the remainder (10.0-20.0%) belonged to the phyla Basidiomycota, Glomeromycota, and Mortierellomycota (1.9-29.6%, 0.1-2.6%, and 0.3-2.7%, respectively) (Fig. 3b). The Ascomycota and Glomeromycota of two treatment groups were a few difference on days 11 and 27, while the Basidiomycota exhibited a great difference on days 11 and 48. At a finer taxonomic level, the abundance of 13 bacterial and 7 fungal taxa (family level) showed significant changes in different compost periods, among which Trichocomaceae, Glomeraceae, Anaerolineaceae, Rhodothermaceae, Limnochordaceae, and Marinilabiaceae were most representative (Fig. S4).
Specific differences of microbiome during composting
The observed differences in α-and β-diversity between the EG and the CK led us to explore more in-depth the differences in taxonomic identity and the abundance of bacterial and fungal taxa. For bacteria, Moheibacter, Halocella, Marinobacter, Petrimonas, and Actinotalea were consistently enriched in the EG, while Halomonas was more abundant in the CK on days 11 and 27 (mesophilic phase) (Welch’s t test, P <0.05, FDR-corrected, Fig. 4a and b). On 48 (thermophilic phase), Methylocaldum, Marinobacter, VadinBC27_wastewater_sludge, Caldicoprobacter, Turicibacter, and Hydrogenispora were more abundant in the EG, while Halomonas and Galbibacter were more abundant in the CK (Welch’s t test, P <0.05, FDR-corrected, Fig. 4c). It was noteworthy that the genera exclusive for the EG and the CK were Marinobacter and Halomonas respectively from day 11 to day 48.
The Marinobacter genus is mostly facultative aerobic heterotrophic and halotolerant bacteria and the main factor favoring Marinobacter abundance was hydrocarbon amendment (Bonin et al. 2015). The genus Halomonas is characteristically halophilic or halotolerant with denitrification function, and can secrete various metabolites (polyhydroxyalkanoates, PHA) with the basic carbon source (Kim et al. 2013). The genera Petrimonas, Actinotalea, and Halocella were mesophilic and able to utilize die-hard substances such as cellulose. By contrast, Hydrogenispora, Caldicoprobacter, and VadinBC27_wastewater_sludge were thermophilic biomass-degrading bacteria and could utilize complex organic compounds (chitin, xylan, and lignin) (Ungkulpasvich et al. 2020; Mhiri et al. 2020; Wang et al. 2021). In terms of metabolic characteristics, they might aid in the degradation of the big molecular substances and refractory organic compounds, and their abundance were strongly modulated by pile environment and positively or negatively affected by temperature. Further experimentation is still needed to decipher the impact of these “enriched” microbes for efficiency and quality of cow dung compost. Moreover, Methylocaldum was obviously enriched (48 d; 0.11% and 3.87%, respectively) in the EG compared to the CK and it could utilize methane, which possibly was contributed to reduce methane emissions during composting (Takeuchi et al. 2014).
For fungi, the genera Pseudallescheria, Melanocarpus, Chaetomium, Coprinellus, and Penicillium showed a higher abundance in the EG than in the CK, while Scopulariopsis was significantly more abundant in the CK on day 11 (Welch’s test, P <0.05, FDR-corrected, Fig. 5a). In the EG, Chaetomium and Penicillium were enriched, however, Microascus was enriched in the CK on day 27 (Welch’s t test, P <0.05, FDR-corrected, Fig. 5b). On day 48, Gamsia, Melanocarpus, Chaetomium, and Penicillium were significantly more abundant in the EG, while Chrysosporium, Scopulariopsis, and Acremonium were enriched in the CK (Welch’s t test, P <0.05, FDR-corrected, Fig. 5c). The Chaetomium and Penicillium are moderate thermophilic and known for its cellulose-degrading capabilities and Melanocarpus is very diverse and cosmopolite fungi and play important roles as decomposers of organic materials (Linkies et al. 2021; Dyer et al. 2014; Feng et al. 2021). They were associated with their ability to decompose complex carbohydrates, thereby contributing to carbon cycling in cow dung compost. Microascus, Acremonium, and Scopulariopsis were more enriched in the CK, which could cause widespread infection was revealed by the researcher. Taken together, the comparative results verified that microbial inoculation contributed to modulate the abundance of specific functional groups and reduce the bacterial pathogens in cow dung compost. The genus level of microbial community was shown in Fig. S5 in detail.
Transcriptional analysis
The transcriptomic analysis of bacterial community in the EG (0 d and 48 d) was generated to investigate genes encoding CAZymes related to the decomposition of cow dung lignocellulose. From six samples (three biological replicates for each period), a total of 2.2 billion cleaned reads (32.2 Gb) were got after filtering and each sample contains approximately 4.1–6.7 Gb (Table. S1). The error rate of transcribe data was 2.4-2.1% and the Q20 and Q30 values exceeded 97.6% and 92.9%, respectively, and reached the basic requirements of gene discovery. A non-redundant transcript cluster was got, including 997,517 unique genes with an average length of 925 bp and an N50 of 1,176 bp, an N90 of 414 bp, respectively. Results indicated that a total of 118,611 genes were up-regulated and 186,660 genes were down-regulated.
Enrichment
230,188 DEGs were carried out an enrichment analysis of GO functions and KEGG pathways. GO function analysis showed that 366, 316, and 423 categories were enriched in biological process (BP), molecular function (MF), and cellular component (CC), respectively. The top 20 GO enrichment circle and GO summary graphs of the DEGs were presented in Fig. 5a and b. Among the top 20 enriched GO entries, the membrane related to cell components possessed higher rich factors. Fifteen items were enriched in BP, among them GO: 0051179 (localization), GO: 0006810 (transport), and GO: 0051234 (establishment of localization) were enriched in more genes (194, 182, and 182, respectively), and the down-regulated genes were also in a higher degree enriched. Lastly, four entries were enriched in MF, among them the P-values of GO: 0015075 (transporter activity), GO: 0022857 (transmembrane transporter activity), and GO: 0016874 (ligase activity) were higher. To further understand the growth status of the microbial community, the DEGs were mapped to the KEGG database. In the top 20 enriched pathways, DEGs mapped to the ribosome (ko03010) occupy the largest proportion, with “RNA degradation (ko03018)” and “Longevity regulating pathway–worm (ko04212)” ranking second and third (Fig. 6c). Combining the analysis results of the GO functions and the KEGG pathway, it was obvious that these pathways involved more in translation, localization, membrane, and biological processes.
There were altogether 39,907 CAZyme-encoding genes were detected in different families, with 18,462 potentially involved in lignocellulose degradation enzymes were found from the auxiliary activities (AA), glycoside hydrolases (GH), and carbohydrate esterases (CE) families. Among them, 2 AAs belonged to the lignin-degrading Enzymes; 14 GHs families belonged to the cellulose-degrading enzymes and 7 CBMs families accessory proteins related to cellulose degradation; 11 GHs and 11 CEs families belonged to the hemicellulose-degrading enzyme system, and 7 CBMs families assisted the catalytic function of the hemicellulase system; 14 GHs families belonged to the cello-oligosaccharides degrading enzyme system and 3 CBMs families belonged to the cello-oligosaccharides degradation enzymes. Using the cluster analysis, we confirmed that the expression levels of lingo-cellulosic enzymes were significantly higher in the thermophilic period compared with the initial period (Fig. S6). These results implied that the inducing mechanism supporting the high expression level of lignocellulose should exist, and it might be intimately connected with the regulation of the microbial inoculation on the resident microbes in cow dung compost.
Relationship among CAZY family genes and microbial community
The correlation matrix among CAZY (Carbohydrate-Active enzymes database) family genes and microbial population in the EG was explored. The complex interactions among bacterial species were performed (Fig. 7). Halomonas and Marinobacter were exclusively represented in the CK and the EG, respectively. Marinobacter exhibited the significantly (p<0.05) positive correlation with VadinBC27_wastewater_sludge, Hydrogenispora, Longispora, Treponema, Methylocaldum, Moheibacter, Limnochorda, Caldicoprobacter, and Tepidimicrobium, with only two significantly (p<0.05) negative connection identified as Halomonas and Turicibacter. By contrast, Halomonas was just the opposite, we found that it showed a significantly (p<0.05) negative correlation with Halocella, Marinobacter, Limnochorda, Turicibacter, Caldicoprobacter, Tepidimicrobium, and Methylocaldum, with only a significantly (p<0.05) positive connection identified as Turicibacter. In the correlation matrix, we observed positive correlations between bacterial populations, which suggested niche overlap, as well as negative correlations, suggesting competition or amensalism (Faust et al. 2012). The significantly enriched bacterial populations in the EG was in general positively correlated, forming well-differentiated clusters. These significantly enriched bacterial populations consisted mainly of resident functional microbes that involved in degradation of complex organic matters.
Fig. 7 showed that the expression levels of cellulose, hemicellulase, and oligosaccharidase genes were significantly (p<0.05) related to Hydrogenispora, VadinBC27_wastewater_sludge, Halomonas, and Methylocaldum. Fig. 8 showed that the expression levels of cellulose, hemicellulase, and oligosaccharidase genes were significantly (p<0.05) related to Chaetomium, Melanoleuca, Pseudallescheria, Penicillium, Gamsia, Pseudogymnoascus, Vishniacozyma, and Aspergillus. These functional microorganisms were more abundant in the EG, which possessed the important feature of microbes associated with degradation of lignocellulose besides. They might be members of the core functional microbiome and most likely better adapted and responded to compost environment in the EG, such as temperature. The diversity and abundance of these microorganisms in EG, as well as their diversity in metabolic traits, makes them potentially important functional microbes in the compost material transformation.