4.2 Effect of fertilization strategies on microbial community structure
All fertilized treatments increased microbial PLFAs (Table 1), showing a fertilizer-induced increase in microbial abundance. Large amounts of N from mineral and organic fertilizers could benefit microbes through a reduction in nutrient competition with rice plants (Jackson, Burger, & Cavagnaro, 2008; Zhu et al., 2018). Moreover, the increase of plant biomass caused by N fertilization would have a direct impact on rhizodeposition and the high availability of labile C to microbes (Christopher &Lal, 2007; Zhu, Vivanco, & Manter, 2016). G− bacteria are usually considered to benefit more from rhizodeposits (Fanin et al., 2019). However, the G+ to G−bacteria ratio was enriched (0.55~0.83 after fertilizers vs. 0.45~0.53 in control) at 0−30 cm soil depths (Figures 2b and c, P <0.05). Moreover, the abundance of G+ bacteria exceeded that of actinomycetes at 0−30 cm. This indicated that both mineral and organic fertilizers preferentially promote G+ bacteria. Pure manure has been demonstrated to induce G−bacteria and reduce G+ bacteria (Peacock et al., 2001). Wang et al. (2014) also found more C incorporation by G−bacteria, which had originally been dominant in the soils. In parallel, they found that litter addition would reduce the G+ to G− bacteria ratio, but its combination with mineral N could buffer this decrease. This indicated that complex substrates induced more G−bacteria and mineral N promoted G+ bacteria, thereby reducing the G+ to G−bacteria ratio. Relatively increased G+ bacteria levels due to mineral N fertilization have also been reported in grassland soils (Denef et al., 2009) and corn fields (Peacock et al., 2001). Consistent with these observations, all treatments with mineral N fertilizers in the present study induced G+ bacteria, even when the soil was predominated by G−bacteria. This indicates a positive response of soil G+ bacteria to mineral N fertilizers.
After partial replacement of inorganic fertilizer with straw, PLFA biomarkers increased at 10–20 cm compared with those under NPK addition alone (Table 2, P <0.05). The fungi to bacteria and G− bacteria ratios were also relatively increased at this soil depth (Figures 1a and d). This indicated that fungi were especially promoted at 10–20 cm, probably owing to more suitable anaerobic conditions for straw decomposition. Simultaneously, qCO2 was observed to be the lowest at this soil depth. Some studies have suggested that higher qCO2 means higher stress or disturbance for microorganisms (Wardle & Ghani, 1995). Under the same level of fertilizer input comparable to straw addition, MBC and PLFAs were largely induced by chicken manure throughout the soil profile (Tables 1 and 2, P < 0.05). Manure addition further increased the G+ to G− bacteria and G+ bacteria to actinomycetes ratios (Figures 1b and c), which may indicate that manure induced more G+ bacteria. In addition, the NPK + OM treatments showed a consistently low qCO2 throughout the 0−40 cm soil profile (Figure 2). Plaza et al. (2004) observed that qCO2 initially decreased after pig manure amendment to soil and then increased with further manure addition. Hence, qCO2 was not negatively correlated with soil fertility or organic fertilizer input. The findings above may preferentially support the connection between low qCO2 and favourable soil conditions for microorganisms.
In addition to its effect on microbial communities through nutrient availability, fertilization can indirectly alter microbial communities through its effect on pH at 0–20 cm soil depths (Figures 3a and b). More organic acids such as oxalic acid (Keiluweit et al., 2015) and acetate (Farrar, Hawes, Jones, & Lindow, 2003) released from rhizodeposits could partly cause a reduction in soil pH, especially in the rooted layers compared with the deeper layers (Table 2, Fig 3a-b,P <0.05). In addition, N fertilizer has been reported to cause soil acidification during N cycling (Bolan, Headley, & White, 1991; Geissler & Scow, 2014). However, in contrast to general assumptions, higher SOC contents were observed in the control compared with those in NPK at 0–10 cm. The soil respiration rate and qCO2 were higher in NPK in this soil layer (Figures 2a and b, P <0.05), and carbon use efficiency was also found to be lower in NPK (Zhran et al., 2020). A greater amount of respired than accumulated C in biomass can partially explain the decrease in SOC. In addition, Loeppmann et al. (2016) found that maize rhizodeposition can decrease the proportional activity of C- to N-cycling enzymes. Mineral NPK fertilizers can increase root biomass and then increase rhizodeposition, but it can also stimulate microorganisms that need both C and N. More C-cycling enzymes will be released by microbes to decompose SOC to maintain a certain microbial C:N ratio (Devi & Yadava, 2006) if microorganisms need more C than rhizodeposits can supply.