BEF-aquaculture

This dataset provides the plot- and replicate-level biomass matrix of hard-substrate fouling communities sampled along a gradient of organic enrichment downstream of offshore fish cages in the Gulf of Castellammare (NW Sicily, Mediterranean Sea). It is the disaggregated data underlying the re-analysis reported in Sarà, Dong and Mangano, "Aquaculture and biodiversity-ecosystem functioning: a path to lay the foundations of ecosystem based approach" (Ecological Applications). The raw field material derives from Sarà et al. (2007, Aquaculture Research 38:1167-1174). Study design: artificial settlement units were deployed for one year, with one upstream undisturbed baseline (CTRL) and post-loss sites at increasing distance downstream from the cages. The sampling hierarchy is Area (CTRL / post-loss) > Site > Plot (3 per site) > Replicate quadrat (3 per plot, 20 x 20 cm), giving 7 sites x 3 plots x 3 replicates = 63 replicate quadrats. For each of 14 taxa (barnacles, bivalves, gastropods, decapods, echinoderm), two measurements per replicate are given: N (abundance, count) and AFDW (ash-free dry weight, g), the biomass proxy used as the ecosystem-function currency. Site-level trophodynamic variables (chlorophyll-a, total suspended matter, particulate organic matter) are also included. The associated study applies the extended Price equation (Fox and Kerr 2012) to partition the change in community biomass between baseline and enriched sites into five additive components (species richness and composition effects of loss and gain, and a context-dependent effect). The data show that total biomass increased at enriched sites, driven mainly by species gains and by enhanced performance of persistent filter-feeders, partly offset by the low per-species contribution of colonizers. Site numbering: the article analyses the four retained post-loss sites, renumbered consecutively PL-1 to PL-4; in the original 2007 numbering these are post-loss sites 1, 2, 4 and 5. Original post-loss site 3 was identified as an outlier and excluded from the analysis, but its raw data are retained here for transparency. Every data sheet carries both the raw and the published site labels. Files: (1) an Excel workbook with sheets README, species_key, biomass_long (tidy format, recommended for re-analysis), biomass_wide, and environmental; (2) a Python script (price_analysis.py) that reads the dataset and reproduces the Price equation decomposition (means, SE, one-sample t-tests, Cohen's d) reported in the article. AFDW cells coded "NA" mean the taxon was present but not weighed (3 cases); N = 0 with AFDW = 0 denotes genuine absence. Units: N per quadrat; AFDW in g; CHL-a and POM in µg l-1; TSM in mg l-1. Requirements: Python >= 3.9 with numpy, pandas, scipy, openpyxl.