The biodiversity of eukaryotes in Bambara groundnut rhizosphere

Bambara groundnut has been observed to resist pest and drought, and still able to produce enormous yield when cultivated on poor soil. The advantages of the crop to farmers includes the fact that it produces enormous yield with very low agricultural input. The aim of the present study was to determine the taxonomic and microbial diversity, and identification of eukaryotic organisms in Bambara groundnut rhizosphere using microbiomeanalyst platform. A total of ten soil samples corresponding to the different growth stages were collected from Bambara groundnut rhizosphere over interval period of 4 weeks at North-West University agricultural farm, Mafikeng campus. These samples were assessed for the presence of eukaryotic organisms through polymerase chain reaction (PCR) of 16S ribosomal ribonucleic acid (rRNA) gene. Metagenomics analysis using culture-independent technique (next generation sequencing (NGS)) by Paired end illumina-Miseq ™ technology sequencing with the prospect of discovering novel eukaryotes with plant growth promoting features was used. Statistical analysis was carried out to profile and confirm identities of detected organisms. Fifty-nine (59) features were detected from the 10 samples by microbiomeanalyst under data normalization and data cleaning. Taxonomic analysis showed that, 69% of the eukaryotes in the samples were Peronosporales while Thalassiosiraceae and others were 30% and 1% respectively. There was profound variance difference in the rhizosphere microbiome mainly at the OTU level which largely attributed to those taxa most strongly depleted by the plant. Thalassiosira pseudonana, which is a centric diatom found in marine environment was observed in this study. This is the first time so far that T. pseudonana is observed in plants’ rhizosphere and its ability to withstand harsh environmental variation might contribute to the ability of Bambara groundnut to be able to withstand drought, pests and diseases.


33
Bambara groundnut, scientifically known as Vigna subterranea, is a self-pollinating annual crop, 34 which was known as Voandzeia Thouars. It is an African indigenous crop that has been grown 35 many years. The crop is more popular in Africa because of its resistance to pests and drought, 36 and its ability to enormous yields when grown on poor soil (Ajilogba, 2019). 37 The rhizosphere is one of the most complex environments with thousands of interactions that 38 play crucial roles in plant's health and it is an ecosystem with several plant-microbe interactions 39 which include symbiotic relationship that leads to the production of nitrogen rich soil. Bambara Soil samples (0.25) g from bambara groundnut rhizosphere were added to the powerbead tubes 62 and gently vortexed to mix the components of the powerbead which should have helped to lyse 63 and disperse the soil particles. Sixty (60) µl of solution C1 was added to the powerbead tube, 64 inverted and vortexed at maximum speed for 10 min. The powerbead tube was then centrifuged 65 at 10,000 x g for 30 s at room temperature. The supernatant is transferred into a 2 ml collection 66 tube provided. Two hundred and fifty (250) µl of solution C2 was added to the solution in the 67 powerbead and vortexed for 5 s after which they were incubated at 4°C for 5 min. The tubes were 68 centrifuged at room temperature for 1 min at 10,000 x g. Six hundred (600) µl of supernatant was 69 transferred to a clean 2 ml collection tube and the pellets were avoided while transferring. Two 70 hundred (200) µl of solution C3 was added, vortexed briefly and incubated at 4°C for 5 min. The 71 tubes were then centrifuged at room temperature for 1 min at 10,000 x g. Up to 750 µl of 72 supernatant were transferred into each clean 2 ml collection Tube. Solution C4 was shaken to 73 mix it before adding 1.2 ml to the supernatant and vortexed for 5 s. Approximately 675 µl of 74 supernatant were loaded onto a spin filter and centrifuged at 10,000 x g for 1 min at room 75 temperature. The flow through was discarded and another additional 675 µl were added and the 76 process was repeated thrice for each sample. Five hundred (500) µl of solution C5 were added to 77 the spin filter and centrifuged at room temperature for 30 s at 10,000 x g. This was to help clean 78 the DNA that was bound to the silica filter membrane and the flow through was discarded from 79 the 2 ml collection tube. The spin filter was then centrifuged at room temperature for 1 min at 80 10,000 x g and placed in a clean 2 ml collection tube. One hundred (100) µl of solution C6 were 81 added to the centre of the white filter membrane to elute the DNA. This was also centrifuged at 82 room temperature for 30 s at 10,000 x g. Finally, the spin filters were discarded and the DNA  defined by clustering at 3% divergence (97% similarity). Finally, these OTUs were taxonomically 100 classified using BLASTn against a curated database derived from RDPII and NCBI 101 (www.ncbi.nlm.nih.gov,http://rdp.cme.msu.edu) and compiled by taxonomic level into both 102 "counts" and "percentage" files. Sequences were considered to be at the species level if they 103 have more than 97% identity to annotated rRNA gene sequenced. They were considered to be 104 at the genus level, family level, order level, class level and phylum level if the sequences have 105 identities between 95 and 97%; between 90 and 95%; between 85 and 90%; between 80 and 106 85%; and those between 77 and 80% respectively (Mills et al., 2012) .  Beta diversity (measures of microbiota differences and similarities among soil samples from 121 growth stages) was also carried out to measure association matrix using Bray-Curtis. Beta 122 diversity analysis was carried out to compare the changes in the presence or absence of 123 thousands of taxa present in a dataset and summarize these into how similar or different two 124 samples. Each sample was compared to other samples generating distance matrix. 125 Clustering analysis was carried out using dendrogram, heatmap and correlation analysis. 126 Clustering analysis was performed with the hierarchial cluster function in the package stat. 127 Correlation analysis was carried out to visualize the overall correlations between different 128 features and used to identify features that were correlated with a feature of interest. 129 Differential abundance analysis was carried out using univariate analysis, metagenome sequence 130 and RNA sequence methods. The univariate analysis was used to identify differently abundant 131 features in microbiome data analysis. Metagenome sequence was carried out using 132 metagenomeSeq R package. The RNA sequence methods was carried out using EdgeR method at 133 OTU level. 134 Biomarker analysis was carried out using Linear Discriminant Analysis (LDA) Effect Size (LEfSe) 135 and random forest (RF). Non-parametric factorial Kruskal-Wallis sum-ranks test was performed 136 to identify features with important differential abundance with regard to class of interest, 137 followed by LDA which calculated the effect size of each differentially abundant features. 138 Random forest analysis was carried out using the random forest package. The outlier measures 139 were based on the proximities during tree construction.

179
The alpha diversity analysis shows the measure across all 10 samples for given diversity index.

222
The results shows features that are considered to be significant based on their adjusted p-value. 223 The results reveals OTU_1956 feature having the highest statistic value of 7.73, while OTU_158 224 feature having the lowest statistic value of 0.39 (Table 2). 225

231
Important features that were identified from the metagenomics sequencing results were 232 emphasized on Table 3 233  (Table 3). However, fitFeature model shaped the count distribution using zero-inflated lognormal 241 model.

243
The results of RNA sequencing showed various (14) features categorized under statistical 244 algorithms for metagenomics datasets. These features are assessed by; log2FC, logCPM, Pvalues, 245 and FDR (Table 4). 246 for performing differential abundance analysis. Both these methods were originally developed 249 for RNA-Seq count data. However, these methods performed better than other statistical 250 algorithms for metagenomics datasets (Table 4) (Table 5). 261  has been observed to withstand environmental conditions of oceans and seas. Also a contrast 312 found was P. infestans that is pathogenic to plants but in this case the effect was not felt in 313 Bambara groundnut because Bambara groundnut is well known to withstand pests and 314 pathogens and also harsh ecological conditions. This is the first time that T. psuedonana would 315 be observed in terrestrial soil rhiszosphere because it is well associated with marine and seas 316 even though not much research has been carried out on T. pseudonana. This study reveals the 317 type of microbial interactions below ground and how it impacts the overall wellbeing and abilities 318 of the plant above ground. It is also an eye opener to the contributions of microbial community 319 to enhancing the ability of Bambara groundnut to grow with yields in harsh environmental 320 conditions. Other research can be carried out to find out what the impacts would be if T. 321 pseudonana is isolated from Bambara groundnut rhizosphere and applied to crops that are not 322 drought tolerant. 323