Reviewer archive: condition-factor analysis

Reviewer archive: condition-factor analysis This folder provides the data and R code for the additional condition-factor analysis requested during peer review of Xiong et al. (2026), "Biodiversity production benefits in aquaculture depend on precision ecological design". Figure R1 compares multispecies and monoculture aquaculture and evaluates the same ecological and operational moderators used in the main analysis. Result at a glance The dataset contains 110 effect-size observations from 38 studies. The overall multilevel meta-analysis gives a pooled log response ratio of 0.013 (SE = 0.014, P = 0.389), indicating no detectable mean difference in condition factor between multispecies and monoculture treatments. Moderator sample sizes vary with data availability: temperature, n = 88; duration, n = 109; density, n = 105; and target-species proportion, n = 108. Categorical estimates based on fewer than 10 observations are exploratory. Files for review Figure_R1_reproducibility.html: rendered code, output, and package versions; this is the recommended first file to inspect. Figure_R1.pdf and Figure_R1.png: final figure outputs. polyculture_dataset_cf.csv: analysis-ready condition-factor data. Figure_R1_reproducibility.Rmd and Figure_R1.R: executable report and equivalent standalone script. Polyculture_CF.Rproj: RStudio project file. Figure R1 contains: (a) study locations, (b) individual and pooled effect sizes, (c) categorical moderator estimates, and (d–g) meta-regressions for temperature, duration, stocking density, and target-species proportion. Data conventions The CSV is UTF-8 encoded; all rows have `Response_type = "condition factor"`. `Water_body` uses `F` for freshwater and `M` for marine. `Group_ID` identifies effect sizes sharing a monoculture control. `Mono_*` and `Poly_*` contain the reported means, standard deviations, and sample sizes. Stocking density is in individuals m^-3. Ecological-position terminology The ecological terms retained in `Position_combination` correspond to the water-column terms used in the manuscript and Figure R1 as follows: Dataset value Manuscript term Figure abbreviation pelagic surface S benthopelagic mid-water M demersal bottom B The `+` separator joins species-level categories: for example, `pelagic+benthopelagic` is `S+M`, and `demersal+demersal` is `B+B`. These are broad comparative groupings rather than exclusive depth restrictions. Reproducing Figure R1 Using R 4.4 or later, open `Polyculture_CF.Rproj` and render the report: rmarkdown::render("Figure_R1_reproducibility.Rmd") Alternatively: source("Figure_R1.R") Both routes read the CSV and overwrite the PDF and PNG. Required packages are: pkgs <- c("cowplot", "dplyr", "ggplot2", "here", "lwgeom", "metafor", "metagear", "patchwork", "rmarkdown", "rnaturalearth", "sf", "stringr", "tidyr" ) install.packages(setdiff(pkgs, rownames(installed.packages()))) The scripts calculate log response ratios, impute unreported standard deviations with the Bracken (1992) method, model covariance from shared controls, and fit multilevel models. Full specifications are in the executable files.