Computational protocol: Cyanobacterial diversity of western European biological soil crusts along a latitudinal gradient

Similar protocols

Protocol publication

[…] 16S PCR amplification and Illumina MiSeq sequencing were used to assess the diversity and community composition of cyanobacteria. Three samples from each site were randomly selected and prepared for molecular analysis. Each sample was split into two, one of which was subjected to a dry treatment where samples were air-dried for 1 day before DNA extraction. The other to a wet treatment where samples were saturated three times a week with H2O and stored in the above-mentioned culture cabinet, under identical conditions, for 2 weeks before DNA extraction. This was intended to increase the likelihood of sampling the entire cyanobacterial diversity found within the samples. DNA was extracted using the PowerSoil DNA Isolation Kit (MOBIO, Carlsbad, CA) according to the protocol. The samples were difficult to lyse; therefore, an additional incubation step was included; the PowerBead tubes were placed in a heat block at 65°C for 30 min followed by 3 × 30 s of bead beating.Fragments of the 16S rRNA gene were amplified by PCR using the cyanobacterial specific primers, CYA359F and an equimolar mixture of CYA781Ra and CYA781Rb (Nübel, Garcia-Pichel and Muyzer ). For each 50 μl PCR reaction, Hotstart-Taq Plus DNA Polymerase Kit (Qiagen, Hilden, Germany) was used containing 1× of 10× PCR Buffer, 200 μM dNTP mix (10 mM of each), 0.4 μM primers (CYA359F and CYA781R) (Eurofins MWG, Ebersberg, Germany), 2.5 units HotStartTaq Plus DNA Polymerase and 100 ng total genomic DNA. The PCR cycling parameters were as follows: initial denaturation for 5 min at 95°C; 35 cycles of denaturing at 95°C for 1 min; annealing at 60°C for 1 min; extension at 72°C for 1 min and the final extension at 72°C for 10 min. Amplicons of the 16S rRNA gene were processed by SeqIT Kaiserslautern using MiSeq sequencer (Illumina) read length 2 × 250 bp for obtaining sequences.Quality control checks were performed on the raw sequence data using FastQC v. 0.10.1. Files were merged, with overlapping paired ends, using FLASH v. 1.2.8 (Fast Length Adjustment of Short Reads). Sequence data were processed using QIIME v. 1.6.0. (Caporaso et al. ) and the QIIME workflow described in Navas-Molina et al. (). For the QIIME quality-filtering process, the default parameters were used, including the operational taxonomic unit (OTU) picking closed-reference pipeline and the identification of chimeric sequences by UCHIME (Edgar et al. ). Sequences were clustered at 97% sequence identity and the taxonomy of the representatives from each OTU was assigned using blast+ (Camacho et al. ) against the Greengenes database (DeSantis et al. ). Sequences were taxonomically classified down to the genus level and OTUs identified other than cyanobacteria were discarded. Samples were rarefied to 10 000 sequences per sample. This method determined the taxonomic composition and genetic variation of the cyanobacteria within the BSCs. Non-metric multidimensional scaling was used to compare sample distances and visualise clustering for the Bray–Curtis dissimilarity using the R phyloseq package (McMurdie and Holmes , version 1.8.2) in R v.3.1.0 (R Core Team ). Non-parametric analyses for multivariate data (adonis) were performed using the R vegan package (Oksanen et al., version 2.0-10). Rarefaction curves were built for alpha diversity measures Chao 1, observed OTUs and Shannon indices calculated by QIIME which suggested that the sequencing effort was sufficient for representing and comparing the cyanobacterial communities. The lowest taxonomic unit was included where possible, and manual taxonomic classification was carried out in order to follow the most recent classification system of cyanobacteria by Komárek et al. (). The exception being Phormidiaceae, which is no longer recognised as a family, but can be included in either the Microcoleaceae or Oscillatoriaceae. Data were compiled for graphical construction in Excel and SigmaPlot v. 10.0. Two-way ANOVA analyses (Statistica v. 10, Stat soft), with Levene's tests for data normality and Fisher LSD post-hoc tests, were performed, on each taxa individually and on the taxa grouped into orders (Nostocales, Oscillatoriales, Synechococcales, Chroococcales), to test for the effects of site and treatment. Data were log transformed after deviations from normality were discovered; in some cases, this rectified the problem. However, for many taxa normality could not be assumed and therefore constituted a deviation from a required ANOVA assumption. However, the interaction between variables was of interest. An ANOVA is considered robust enough if the P values are highly significant and can be confirmed by a non-parametric equivalent, such as a Kruskal–Wallis test (Fry ; Zar ). In cases where log transformation, highly significant P values nor Kruskal–Wallis tests supported the data the taxa were omitted from the statistical analysis. See Table S1 in the supplementary material for results of Levene's and Kruskal Wallis tests that lend support to the ANOVA tests demonstrated in the results.Fastq files containing the raw data from this study were submitted to the NCBI sequence read archive (www.ncbi.nlm.nih.gov/sra) and can be accessed by the accession number PRJNA325717. […]

Pipeline specifications

Software tools FastQC, QIIME, UCHIME, phyloseq, SigmaPlot, Statistica
Databases SRA
Applications Miscellaneous, 16S rRNA-seq analysis
Diseases Carotid Artery Thrombosis
Chemicals Nitrogen