Computational protocol: Trait and phylogenetic diversity provide insights into community assembly of reef‐associated shrimps (Palaemonidae) at different spatial scales across the Chagos Archipelago

Similar protocols

Protocol publication

[…] Based on a previous Palaemonidae phylogenetic study (Kou et al., ), we used four genes to construct a focused community phylogeny; partial fragments of the 16S ribosomal RNA (rRNA) gene (~368 bp), and partial fragments of three nuclear genes; enolase (~405 bp), PEPCK (~521 bp), and NaK (~620 bp). Nineteen of the twenty species from the metacommunity were represented by at least two genes in the consensus phylogeny (see Table ). Only Exoclimenella maldevensis was not included in the consensus phylogeny as we were only able to amplify the 16S gene for this species. As this species was rare in the community, occurring only once, it was excluded from further analysis. An additional 26 species were included in the phylogeny (from Chagos samples and available specimens on GenBank) to provide more information on the evolutionary relationships between species in the metacommunity. Phylogenetic trees were constructed under Bayesian Inference (BI) analysis in MrBayes v.3.2 (Ronquist et al., ) (see Table for models of evolution used), on the online CIPRES Science Gateway (Miller & Schwartz, ) for the consensus alignment and for each gene tree separately. A composite metacommunity phylogeny was produced in APE using the phylogeny (Figure ; Paradis, Claude, & Strimmer, ). All sequences were catalogued on GenBank. See Appendix for a detailed methodology. [...] To test for phylogenetic signal in the quantitative traits, we used the K‐statistic (Blomberg, Garland, & Ives, ), using the R package “phyltools” (Revell, ). This was preformed twice; once incorporating sampling error following Ives, Midford, and Garland (), as our data have within‐species variation which is not accounted for in other methods to the best of our knowledge, and for a second time without taking within species variation into account. Phylogenetic signal in the nominal trait of habitat association was tested using Maddison and Slatkin () method which compares the minimum number of trait changes to the distribution of trait changes drawn from a null model. We used function “phylo.signal.disc” in R environment, developed by Enrico Rezende (Universidad Autònoma de Barcelona) (http://grokbase.com/k-for-discrete-unordered-traits). To investigate functional (trait) diversity and phylogenetic patterns in trait diversity across distinct spatial scales, we used the third proposition of Pavoine, Marcon, Ricotta, and Kembel () to divide Rao's measure of diversity, named quadratic entropy (QE), (Rao, ) across spatial scale, using the R package “adiv” (Pavoine, ). This partitioning of diversity is adapted to unbalanced sampling design. Quadratic entropy is also relatively robust to sampling‐effects because this method is an extension of the Simpson index for functional and phylogenetic data, which gives high weight to common species (Lande, ). Therefore, rare species perhaps not identified by under‐sampling are unlikely to impact the index even if they are functionally or phylogenetically distinct from others. The QE index of diversity uses the phylogenetic tree, distributions of relative abundances of species in a community, and a matrix of trait distances among species obtained by Gower distance (Gower, ), to assess whether there is any phylogenetic and/or trait clustering in the metacommunities and local communities (Pavoine et al., ). Phylogenetic and/or trait clustering are measured using the beta diversity standardized effect size (SES), which calculates the observed beta diversity minus the mean of simulated beta diversities, divided by the standard deviation of simulated beta diversities. The trait‐based apportionment of quadratic entropy across spatial scales will be referred to as the trait quadratic entropy test (TQE) and that based on phylogenetic data as the phylogenetic quadratic entropy test (PQE). We measured beta diversity, using TQE and PQE, at three distinct spatial scales (1) among atolls, (2) among reefs within atolls, and (3) among coral colonies within reefs, to determine whether there was spatial structure to the community. To investigate how robust the community phylogenetic diversity patterns are to the evolutionary information used, we ran the apportionment of diversity (PQE test) on each gene tree separately in addition to the consensus phylogeny. […]

Pipeline specifications

Software tools MrBayes, CIPRES Science Gateway, APE, PHYSIG
Application Phylogenetics
Organisms Homo sapiens