Gelcompar 2 5

Fingerprinting of microbial communities using single-stranded conformational polymorphism was conducted as described by Schwieger and Tebbe (1998 ). Amplification of the nifH gene fragment was performed using a nested PCR approach with primer pairs nifH4/nifH3 (Zani et al.2000 ) and nifH11/nifH22^P (Yeager et al.2004 ). The obtained amplicons were separated and analyzed according to Bragina et al. (2012 (link)). Comparisons of SSCP generated nifH community profiles were performed using GelCompar II 5.1 (Applied Maths, Kortrijk, Belgium). The cluster analysis was performed with the following settings: dendrogram type: unweighted pair group method with arithmetic mean (UPGMA); similarity coefficient: curve based: Pearson correlation; position tolerances: optimization: 0%, position tolerance: 1%. Based on the Pearson similarity matrix, a multidimensional scaling (MDS) ordination plot was constructed. Pearson correlation matrices were additionally subjected to significance tests of pair-wise similarities by applying permutation analyses (p ≤ 0.05) using the permtest package of R statistics 2.13.1 (The R Foundation for Statistical Computing, Vienna, Austria) with 10⁵ random permutations of sample elements (Kropf et al.2004 ; R Development Core Team 2011 ).

Fingerprinting by SSCP analysis was carried out as described by Schwieger & Tebbe38 (link). Bacterial 16S rRNA gene sequences were PCR-amplified using the eubacterial primer pair Unibac-II-515f and Unibac-II-927r^P. Separation and analysis were performed according to Köberl et al.39 (link). Comparisons of generated bacterial community profiles were performed using GelCompar II 5.1 (Applied Maths, Kortrijk, Belgium). Cluster analyses were performed with the following settings: dendrogram type: unweighted pair group method with arithmetic mean (UPGMA); similarity coefficient: curve based: Pearson correlation; position tolerances: optimisation: 0.2%, position tolerance: 1%. Multidimensional scaling (MDS) ordination plots were constructed based on the Pearson similarity matrices. These matrices were additionally subjected to significance tests of pair-wise similarities by applying permutation analyses (p ≤ 0.01) using the permtest package of R statistics 3.2.0 (The R Foundation for Statistical Computing, Vienna, Austria) with 10⁵ random permutations of sample elements40 (link)41 . Excised and re-amplified DNA fragments were sequenced at LGC Genomics (Berlin, Germany).