4 DISCUSSION
Intensified livestock production system contributes significantly to environmental impacts including water, air and soil quality by altering the biogeochemical cycling of carbon, nitrogen and phosphorus (Bai et al., 2018; Leip et al., 2015; Pelletier & Tyedmers, 2010). In this study, higher amount of nitrate was detected along the soil profile applied with cattle manure compared with the adjacent non-manured woodland, indicating higher risks of nitrate pollution even down to the subsurface soils with continued manure application. This result was in good agreement with a meta-analysis based on over 7000 cropland samples, which showed large accumulations of nitrate in the soil deep down to 4 meters (Zhou et al., 2016). Soil δ15N-NO3- value further corroborated the high proportion of manure-derived nitrate in the maize field, which is consistent with previous research working on nitrate contamination sources in various environments (Fenech et al., 2012). While some researchers mentioned that the effect of manure addition on nitrate leaching varied with soil texture, climatic condition as well as the applied amounts (Gai et al., 2019; Maeda et al., 2003). Over-use of manure results in nitrate redundancy that gives rise to environmental concerns (Herrero & Thornton, 2013), e.g., groundwater contamination, soil nitrogen losses. Similar result was also obtained from a long-term field study carried out in Nebraska in USA applied with different rates of manure application (Nguyen et al., 2013). However, in the above mentioned study, they suggested that manure if applied below or at the rates equal to crop N requirements has many benefits for food production and environmental protection. Recently, the possibility of substituting manure for commercial fertilizer has been addressed based on 141 published studies and concluded that recycling of livestock manure in agroecosystems would reduce nitrogen losses and increase food production especially in upland soils (Xia et al., 2017). Critical land management in crop production is therefore necessary in keeping sustainability of food production and agricultural systems.
Soil microbial diversity exhibited distinct response to the environment disturbance (Falkowski et al., 2008). In this study, manure addition had no significant impact on bacterial OTU richness and Shannon diversity in the maize field compared to woodland for both seasons, indicating that the change of land use in this site had no significant impact on bacterial α diversity. Generally, agricultural soil microbial diversity tend to be spatially homogeneous when compared to natural habitats because of the interference from human activities (Kennedy & Smith, 1995). Soil bacterial diversity showed a decreasing trend from woodland to cropland (Zheng et al., 2017), which is contrast to our observation. Previous research also observed that soil bacterial diversity responds strongly to the land use change of Amazon rainforest and resulted in biotic homogenization in agricultural soils (Rodrigues et al., 2013). This discrepancy regarding the little difference of bacterial diversity between these two land uses was probably attributed to the addition of manure in the maize field. Additionally, not all the change of land uses would alter bacterial α diversity (Jesus et al., 2009). For example, soil bacterial diversity kept constant following the conversion of cropland to orchard in a degraded karst system (Liao et al., 2018). Our results suggested that the conversion of woodland to cropland did not significantly change bacterial α diversity.
Contrasting to the result of bacterial OTU richness between these two land uses, clear separation of bacterial structure between the maize field and woodland was detected even to the deep layers, which was further confirmed by PerMANOVA analysis. This difference may be explained by the following reasons. Firstly, manure contains numerous microbial resources including liable C and N, providing soil nutrients for microbial growth, especially for the bacteria (Enwall et al., 2007; Sun et al., 2004). Secondly, manure application will change soil physical and chemical properties by altering soil aggregate stability, water and gas exchange, which in turn improving microbial living conditions (Guo et al., 2019). Additionally, manure harbors a large amount of microorganisms (Meng et al., 2019), and some of which will be introduced and persisted in soils from several days to several months depending on soil types and manure sources (Udikovic-Kolic et al., 2014). All these indicated that manure application favor microbial growth, enabling a shift in soil bacterial community composition (Peacock et al., 2001; Shen et al., 2010). The differences between these two land uses were more apparent in the network analysis as maize field networks harbor more modular than woodland. The co-occurrence network in woodland harbored a denser and highly connected bacterial community compared with that in the maize field, nevertheless, the bacterial community in maize field is more specialized with different ecosystem function.
Most interestingly, the shift of bacterial composition kept consistent along the soil profile, as site-specific pattern of bacterial community structure was detected in this study. Differences in spatial distribution of soil bacterial community was largely affected by the vertical gradients of soil properties (Jiao et al., 2018). The increase of ecological heterogeneity with increasing depth in our studied site may be attributable to the significant response of bacterial β diversity along the soil profile. Similar to our study, researchers in four Alaska soil cores also indicated that microbial assembly process was mainly determined by environmental factors rather than depth (Tripathi et al., 2018). In both land uses, the ses.MNTD mean values deviated significantly from zero, indicating that bacterial assemblages had higher phylogenetic clustering than expected by chance. This finding is consistent with other previous studies (Horner-Devine & Bohannan, 2006; Stegen et al., 2012), which indicated that soil bacterial community were more structured by environmental filtering in a wide range of environments. Significant correlation between ses.MNTD and depth further suggested that bacterial community are shifted from more phylogenetically clustered assembly to less along soil depth. Similar results have been found in various habitats, like Alaskan soil cores (Tripathi et al., 2018), Tibetan Plateau (Chu et al., 2016), which showed that environmental filtering are more important in the assembly of bacterial community in surface soils than in deeper soils.
In this study, we found higher concentrations of nitrate in the maize field with long-term applications of livestock waste. This raises the question about which bacterial species responded mostly promptly or persisted longer following manure addition in the context of high nitrate scenario. Pearson correlation analysis between the dominant bacterial groups and nitrate content suggested that nitrate was significantly correlated with phyla of Bacteroidetes (P< 0.001), Nitrospirae (P < 0.01),Firmicutes (P < 0.001), andGemmatimonadetes (P < 0.001) in the maize field (Table S3), but not correlated with special groups in the woodland. Furthermore, many bacterial groups at phylum or order levels showed significant differences in the two land uses in the spring than in the autumn, especially Bacteroidetes . Accumulating evidence showed that Bacteroidetes is widely present in a range of habitats including plant (Thomas et al., 2011), soil (Lauber et al., 2009) and animal gut (Tajima et al., 1999). Their mammal origin might explain the higher abundance and sensitive response in the maize field especially in spring. Our result is consistent with a previous study by Wolinska and his colleagues (Wolinska et al., 2017), who suggested that Bacteroidetes could be recommended as a sensitive indicator for agricultural soil type. Additionally, Bacteroidetes is specialized in decomposing high molecular weight compounds such as polycyclic aromatic hydrocarbons and xenobiotics (Fernandez-Gomez et al., 2013), which could be explained by its copiotrophy life strategy exhibiting more responsive to rich available nutrients (Ho et al., 2017; Pepe-Ranney et al., 2016). The enrichment of this microbial group in response to the addition of exogenous manure, suggests their ability in processing complex organic matter in soils and therefore improving soil fertility.
CONCLUSIONS
Significant changes in soil nutrient availability with the addition of manure contributed to the separation of soil bacterial community structure between different land uses in our study. More sensitive bacterial groups were detected in response to the manure addition especially in the beginning. Bacteroidetes phylum was one of the most responsive groups to the addition of manure addition, which could be a potential bio-indicator for agricultural usage. These results provide solid and necessary information for a better understanding of the succession characteristics of bacterial diversity as well as the driving factors in the maize field receiving cattle manure over a long term. Bacterial diversity is beneficial to terrestrial ecosystem in increasing disturbance tolerance and maintaining ecosystem services. Network analysis further highlights strong differences in network structure between the maize field and woodland that is more related to ecosystem function. We propose that future research should focus on the microbial diversity and function in assessing environmental risks of manure disposal as well as strategies to reduce nitrate leaching.