Objective  A slow-release fertilizer was applied on a cassava lot to analyze the responses of the fungal community and C, N, P, and S functional genes in the rhizosphere soil.

  Method  A field experiment was conducted with treatments of no fertilization (T1), basal application of double-coating slow-release fertilizer C2 (T2), and C2 applied 34 d after planting (DAP) (T3). Rhizosphere and bulk soil samples were collected at 77, 104, and 134 DAP to determine fungal diversity according to ITS rRNA sequenced by a high-throughput Illumina Miseq PE300, copies of 72 functional genes of C, N, P, and S cycles (including total DNA) by the gene chip technology, and available nutrients by chemical analysis for a correlation analysis.

  Result  (1) Significant differences on the relative abundance (RA) of Mortierellomycetes, Tremellomycetes, and Orbiliomycetes were found in the rhizosphere soils on 104 DAP showing T2＜T1, while that of Scatterocysts on 134 DAP indicating T2＞T1. Fungal class Scatterycetes under T2 on 134 DAP, Rubiaceae under T3 on 77 DAP, and Coprochetes under T3 on 104 DAP were relatively enriched in the rhizosphere than in the bulk soil. The RAs of the rhizosphere fungi also differed significantly on time of sampling and under different treatments. They were 134 DAP＜104 DAP for Mortieromycetes under T1, 134 DAP＞77 DAP for Scatterocystae and Coprochestae under T2, and 104 DAP＞77 DAP for Sphaeromycetes under T3. (2) The Sobs, ACE, and Chao1 indexes under T1 on 104 DAP, T2 on 104 DAP, and T3 on 134 DAP were significantly higher in the rhizosphere than in the bulk soil (P＜0.05 or P＜0.01). The Shannon index of rhizosphere soil was lower under T1 than under T2 or T3 on 77 DAP. Under T1, the index was 104 DAP＞77 DAP; and under T2, it was the opposite. The Simpson indexes ranked in the order of the rhizosphere soil under T1 on 77 DAP＞the bulk soil under T1 on 77 DAP＞the rhizosphere soil under T1 on 104 DAP＞the rhizosphere soil under T2 on 77 DAP＞the rhizosphere soil under T3 on 77 DAP. (3) The LEfSe analysis indicated the fertilizer applications enriched one class, one order, and two families of fungi in the rhizosphere soil, whereas the bulk soil was more abundant in two species on 77 DAP, in 3 orders, one family, and one genus on 104 DAP, and in one phylum, one order, one family, and one genus on 134 DAP. Two orders on 104 DAP and one class on 134 DAP were enriched in the rhizosphere soil. (4) On 134 DAP, the 9 functional genes, such as lig in the bulk soil under T1, had significantly more copies than in the rhizosphere soil. In the rhizosphere soil, the RAs of chiA and aclB under T1 on 77 DAP were higher than those on 104 or 134 DAP. (5) AK significantly correlated with 31 functional genes on 104 DAP. Some fungal classes, such as Tremella, Sarcoidales, Claviculaceae, and Sphaeromycetes, significantly correlated with 40, 15, 14, and 9 other functional genes, respectively.

  Conclusion   Fertilization by ways of T2 or T3 enriched the diversity and abundance of cassava rhizosphere fungal community. Fertilizer used, application time, and rhizosphere could all significantly affect the fungal community structure and some functional genes in the soil. The correlations might lead to further studies to unveil the intricate ecosystem.