Fig 1.
Study sites in Germany (A), experimental design of paired agroforestry and conventional monoculture cropland (B), and a picture taken at the tree-crop interface of the agroforestry cropland system at Dornburg (C). The soil types at the three study sites were Calcaric Phaeozem soil (near Dornburg, Thuringia), Gleyic Cambisol soil (near Forst, Brandenburg), and Vertic Cambisol soil (near Wendhausen, Lower Saxony). Photo taken by G. Shao.
Table 1.
Site characteristics and management at the three study sites of paired temperate agroforestry and monoculture cropland.
Fig 2.
Kernel density estimation of the quantification cycle (Cq) values for non-normalised and normalised samples (prior to library preparation) (A), prepared libraries normalized by photo densitometry (B), and the distribution of the raw sequence reads per library (C). Individual Cq values are plotted as marks below the density curves for non-normalised and above for normalised samples; curves were smoothened using a bandwidth of 0.3 (A). The box indicates the q25, q50, and q75, whiskers range from q25 or q75 to 1.5 × the interquartile range, dots represent individual data points (C). n = 60 samples.
Fig 3.
16S rRNA gene abundance of soil bacteria (A) and relative abundance of taxonomic groups of soil bacteria at class level (B) in paired temperate agroforestry and monoculture cropland systems in three different soil types. Samples in the agroforestry systems were collected in the tree row as well as at 1 m, 7 m, and 24 m distance from the tree row within the agroforestry crop row. Bars represent means with standard deviation (n = 4 per soil type × sampling location) (A). Different uppercase letters indicate significant differences among the sampling locations (one-way analysis of variance with Tukey’s honestly significant difference test or Kruskal–Wallis test with multiple comparison extension) (A). Soil bacterial classes with an overall relative abundance of < 0.5% were considered as rare classes and merged with unassigned classes (B).
Fig 4.
Alpha diversity measures (Shannon index (H’) (A) and Pielou’s evenness index (J’) (B)), non-metric multidimensional scaling (NMDS) of Bray-Curtis dissimilarities (C), and Bray-Curtis similarity network (D) of soil bacterial ASVs in paired temperate agroforestry and monoculture cropland systems in three different soil types. Samples in the agroforestry systems were collected in the tree row as well as at 1 m, 7 m, and 24 m distance from the tree row within the agroforestry crop row. Coloured vertical bars represent the standard deviation, coloured horizontal bars the mean (A, B). Coloured shapes represent individual data points (A, B). Different uppercase letters indicate statistically significant differences at p < 0.05 (one-way analysis of variance with Tukey’s honestly significant difference test or Kruskal–Wallis test with multiple comparison extension) (A, B). Solid lines in the NMDS span from the centroid of each group to the individual data points (C). Edges in the similarity network were drawn between samples with a Bray-Curtis similarity ≥ 0.75 (D). The network was spatially arranged applying the Fruchterman-Reingold algorithm (D). AF = agroforestry system, ASV = amplicon sequence variant.
Fig 5.
Z-score normalized relative abundance (A) and relative abundance (B) of selected taxonomic soil bacterial genera in paired temperate agroforestry and monoculture cropland systems in three different soil types. Samples in the agroforestry systems were collected in the tree row as well as at 1 m, 7 m, and 24 m distance from the tree row within the agroforestry crop row. Horizontal bars represent the mean relative abundance, coloured shapes represent individual data points (B).
Fig 6.
Abundance of N fixation gene nifH in paired temperate agroforestry and monoculture cropland systems in three different soil types.
Samples in the agroforestry systems were collected in the tree row as well as at 1 m, 7 m, and 24 m distance from the tree row within the agroforestry crop row. Bars represent means with standard deviation (n = 4 per soil type × sampling location). Different uppercase letters indicate significant differences among the sampling locations within the same soil type (one-way analysis of variance with Tukey’s honestly significant difference test or Kruskal–Wallis test with multiple comparison extension).