Computational protocol: Tree species, tree genotypes and tree genotypic diversity levels affect microbe-mediated soil ecosystem functions in a subtropical forest

Similar protocols

Protocol publication

[…] The effects of tree species identity, tree genotype and genotypic diversity level on overall soil enzyme patterns for the six enzymes were analyzed using Permutational Multivariate Analysis of Variance (PERMANOVA) based on Euclidean distances implemented in PAST. Statistical significances were based on 999 permutations and Bonferroni-corrected P values were applied in all cases when more than two groups were compared with PERMANOVA. Non-metric multidimensional scaling (NMDS) ordination based on Euclidean distances with the envfit function of the vegan package in R was used to investigate and visualize correlations among different factors influencing overall soil enzyme activities. All significant factors (P < 0.05) were plotted in the respective NMDS ordinations. We determine the individual factors that were most influential for the overall pattern of soil enzyme activity using distance-based redundancy analysis (dbRDA), based on Euclidean and Bray–Curtis distances with the function capscale in vegan. We tested for collinearity between all factors using Spearman’s rank correlation (ρ) before carrying out dbRDA. When two or more factors were correlated (ρ > 0.70 or ρ < −0.70 and P < 0.01), only one factor was retained for further analysis. In this case, only total soil C and total soil N were highly correlated (ρ = 0.86, P < 0.01), thus only N was kept for dbRDA. Four factors including tree species identity, genotypic diversity level, soil pH and soil water content were significant based on dbRDA and were selected for variance partitioning analysis. Variance partitioning analysis was performed using the varpart function in the vegan package in R to take into account the influence of four significant factors based on dbRDA (including both soil and plant related factors) on overall enzyme activity pattern. We used the same enzyme activity matrix and factors as used for NMDS analysis. We also test whether shared fraction of these four factors explains the variance of soil enzyme activity. The effects of tree species identity on specific enzyme activities were analyzed (as between plot factor) using split-plot ANOVA simultaneously accounting for tree genotype (within plot factor) and genotypic diversity level (between plot factor) (IBM SPSS Statistics 22, New York, NY, USA). Effects of tree genotype (within plot factor) and genotypic diversity level (between plot factors) were also analyzed for each tree species using a split-plot ANOVA. All data sets were tested for normality and the equality of group variances using a JB test and the Levene statistic and log10 transformed when necessary. Fisher’s Least Significant Difference (LSD) post hoc test was applied when more than two treatments were tested (IBM SPSS Statistics 22, New York, NY, USA). […]

Pipeline specifications

Software tools vegan, SPSS
Application Miscellaneous