Computational protocol: Dissection of niche competition between introduced and indigenous arbuscular mycorrhizal fungi with respect to soybean yield responses

Similar protocols

Protocol publication

[…] The first PCR products were subjected to the 2nd PCR to attach dual indices and Illumina sequencing adaptors using Nextera XT Index Kit v2 (Illumina, Tokyo) according to the 16 S Metagenomic Sequencing Library Preparation protocol (https://support.illumina.com/downloads/16s_metagenomic_sequencing_library_preparation.html). The libraries were pooled, adjusted to 4 nM DNA, denatured with NaOH, and then diluted with the hybridization buffer to the final concentration of 6 pM DNA, and 300-bp paired-end sequencing was carried out by the Illumina MiSeq platform using MiSeq Reagent Kit v3 (600 Cycles) (Illumina K.K., Tokyo).The nucleotides with a quality value (QV) <30 in the 3′ terminal and the adapter-index sequence in the 5′ terminal were trimmed from the MiSeq reads by PRINSEQ v0.20.4, and those shorter than 200 bp were excluded. After the quality filtering, an overlap fragment of the 300-bp-paired reads (read 1 and read 2) were constructed by using COPE v1.1.3 with the minimum and maximum overlap lengths of 10 and 300 nt, respectively. The merged sequence reads were subjected to BLASTN searches against reference sequences (database) composed of 412 operational taxonomic units (OTUs) of glomeromycotinan (AM) fungi defined in this study (Supplementary Methods , Table , and Fig. ) and 82,208 sequences of non-glomeromycotinan fungi obtained from Ribosomal Database Project, in which sequence reads similar to an AM fungal OTU and those similar to a non-glomeromycotinan species were assigned to the OTU/species with different criteria as follows. For the reads similar to AM fungal OTUs, only those that met the criteria of E-value ≤ −100, ≥95% nt identity, and alignment length ≥330 bp with one of the AM fungal OTUs were assigned to the OTU. Whereas, for the reads similar to non-glomeromycotinan sequences, only those that met the criteria of E-value ≤ −100, ≥95% nt identity, and alignment length ≥220 bp with one of the 82,208 sequences were assigned to the species. All these analyses were executed in the open web interface “Arbuscular mycorrhizal fungi classification pipeline” (http://amfungi.kazusa.or.jp) constructed in this study. [...] All statistical analyses were performed with R 3.2.3. Two-way analysis of variance (ANOVA) with random effects for block differences were applied to evaluate the effects of the inoculation and P fertilizer application on shoot P concentration and grain yield. Vegan package for R was employed for β-diversity analysis with Bray-Curtis similarity index (read-abundance data based index) and for permutation multivariate analysis of variance (PERMANOVA) in which Bray-Curtis index was used as a measure of similarity (9999 permutations). In regression analyses all OTUs obtained by sequencing of the inoculum fungus R-10 were combined as ‘R-10-type OTUs’, in which ‘read abundance of the R-10-type OTUs’ represents the sum of the sequence reads assigned to the OTUs. Simple linear regression models were applied to analyze correlations between read abundances of the R-10-type OTUs (log-transformed) and MPNs of indigenous AM fungal propagule (log-transformed) and between the read abundances and soybean yield responses to the inoculation. The yield responses were calculated by the equation () as follows:1Yieldresponsetoinoculation=(Yieldininoculatedplot−Meanyieldincontrolplots)/MeanyieldincontrolplotsThe multiple linear regression model was applied to evaluate to the effects of the read abundance of R-10-type OTUs and the environmental factors (Supplementary Table ) on the yield responses to the inoculation. The best model was selected with reference to Akaike’s information criterion (AIC) calculated by the stepwise method using MASS package in R. The logistic regression model with random effects for block differences was applied to evaluate the effect of environmental factors on the read abundance of R-10-type OTUs using glmmML package in R. The best model was selected with reference to AIC calculated using MuMIn package.To dissect competition between the introduced (inoculum) and indigenous AM fungi, the following two indices were employed. ‘Commonness’ that is the index referred to as niche breadth in Levins (2013) and Pandit et al. was employed to evaluate their habitat specialization and calculated for each indigenous OTU by the equations ( and ) as follows:2Commonness=1/∑(i=0)nPij23Pij=MeanreadnumberofOTUjintrialiSumofmeanreadnumbersofOTUjineachtiralwhere only the read number data in the uninoculated control were used to compute. OTUs with a higher value of commonness distribute more evenly across the habitats and thus are considered to be habitat generalists. Rare OTUs of which the mean relative abundance was less than 0.2% of total read were not considered in this analysis. Robustness of each indigenous OTU against the introduction of R-10 fungus was defined as a ratio of the mean read abundance in the inoculated plots to that in the control plots in each trial. […]

Pipeline specifications

Software tools PRINSEQ, BLASTN, vegan, MuMIn
Applications Phylogenetics, Metagenomic sequencing analysis
Organisms Glycine max, Fungi
Chemicals Phosphorus