Extended Data Table 1 Results of statistical tests for environmental parameters quantified in OMCO collector tracks, plume area and control site

From: Long-term impact and biological recovery in a deep-sea mining track

  1. NS = not significant (p > 0.05). NA = Not applicable. Generalized linear (GLM) statistical models were independently developed using R (version 4.4.1) to examine whether the variable under investigation (e.g. faunal density) could be explained using the explanatory variable of disturbance type (categorical, e.g. track, plume and control). Quasi-Poisson errors were used for count variables, where variance appeared to be related to the mean. Gaussian errors were used for continuous data and selected after verifying normality (by visual examination of quantile-quantile plot) and homogeneity of variance (by visual examination of conditional boxplots). Kruskal-Wallis was used where variables were not clearly normally distributed. Numbers of observations (n) are indicated in the comparisons column. Note very low replication (n = 2) in microbial carbon uptake and biomass, where tests should be treated with caution. All measurements were made from distinct samples.
Back to article page