Biodiversity patterns of epipelagic copepods in the South Pacific Ocean: Strengths and limitations of current data bases

in R software (R Core Team, 2021). “Deviation” of temperature is defined as the mean standard deviation between time slices over time elapsed, which was calculated at the beginning and end of each of the two years temporal lags and dividing this result by the length of the interval to quantify deviation over time for each time slice, to finally average the result across all time slices. Later, “sta- bility” was obtained when taking the inverse of the deviation and scaling it between 0 and 1 [50]. Therefore, a total of seven variables were included in our analyses and are available in Zenodo repository: https://doi.org/10.5281/zenodo.8257222 (version 1). These environmental variables were preprocessed by standardizing them subtracting their mean and dividing them by their stan- dard deviation for later performing variable selection by using Random Forests through the pack- age VSURF [51] implemented in R software (R Core Team, 2021). This is a two-stage strategy based on a preliminary ranking of the explanatory variables using the random forests permuta- tion-based score of importance [52]. We selected a subset of important variables from the second or the “interpretation” step according to the reduction of the out-of-bag error by adding them into the decision training models. These variables were later used for analysis of predictive models of species richness and species composition. Further, for the selected environmental variables, we calculated the Spearman rank-order correlation coefficient matrix to visualize their degree of asso-

ciation as well as check for redundancy. Spatial analyses To analyze the species composition and its spatial distribution, the most dominant species (most recurrent) were extracted from the data bases and then associated with major current systems of the SPO, assuming they were representing distinctive species assemblages from these currents. Also, for beta diversity and its components, turnover and nestedness, a pres- ence-absence matrix (PAM) was generated through the package fuzzySim in R software (R Core Team, 2021) [53] from the list of species that occurred for each range of values. This PAM was later matrix plotted in PAST software [54], from which we selected visually the dom- inant species, i.e., the ones occurring in all ranges of values for each component. Biodiversity hotspots were determined based on spatial-statistical criteria to describe biodiver- sity patterns for the 0–200 m depth water column. It was specifically aimed to detect cells or groups of cells (i.e., spatial clusters) with species richness greater than the richness expected in a random distribution. First, the evaluation of whether copepods depict spatial autocorrelation using Moran’s I statistic was conducted. Second, spatial hotspots were defined in ArcMap 10.4.1 (ESRI 2016) software using the High/Low Clustering (Getis-Ord G*) statistic [55], which identi- fies spatial concentrations of an entity (in this case species richness per cell) or areas that contain higher/lower values than expected by chance for a given study area. For a hot spot, the observed

G* index indicated high values of richness clustered in the study area; whereas for a cold spot, the observed G* index indicated that low values of richness were clustered in the study area. Species richness (alpha diversity) is the absolute number of species living in a given area, giving equal weight to all resident species [56]. Species composition (namely beta diversity) can be viewed as a measure that compares inventory diversity at two different scales (alpha and gamma diversity). Beta diversity is also a measure of similarity between sites [57], and thus reflects the differences between local biological communities within a region. These differences arise from two phenomena which should be distinguished when understanding how biological diversity is distributed: nestedness and spatial turnover [58]. Nestedness of species occurs when the biota of sites with smaller numbers of species are subsets of the biota at richer sites. Contrary to nestedness, spatial turnover implies the replacement of some species by others as a consequence of environmental sorting or historical and spatial constraints [59]. An application of knowing these components of beta diversity is the attempt to use differences between PLOS ONE Biodiversity of copepods in the South Pacific PLOS ONE | https://doi.org/10.1371/journal.pone.0306440 July 11, 2024 5 / 28