Through murky waters

Citations and original sources

Paper. Chong, F., Long, M., De Palmas, S., Srenh, S., Rog, S. M., Glue, M., Vibol, O., & Reimer, J. D. (2026). Through murky waters: conservation implications from evaluating the drivers of diversity and abundance on coral reef communities in Cambodia. Biodiversity and Conservation, 35(11), article 251. doi:10.1007/s10531-026-03441-6

Dataset. Chong, F., Long, M., De Palmas, S., Srenh, S., Rog, S., Glue, M., Vibol, O., & Reimer, J. (2026). Data and code used in “Through murky waters: conservation implications from evaluating the drivers of diversity and abundance on coral reef communities in Cambodia” [Dataset and R code; deposited record 21910583]. Zenodo. doi:10.5281/zenodo.21910583

Paper and dataset published 26 August 2026. The dataset citation identifies the deposited version used here. Both are shared under Creative Commons Attribution 4.0.

Read the paper · Paper PDF · Published supplementary data and code · Dataset repository

Survey and findings

Each reef was visited once. Four 20 m transects at approximately 5 m depth supported fish and invertebrate counts in 5 m-wide belts; benthic points were recorded every 25 cm. Field dates run from 8 May 2021 to 10 July 2023.

The paper reports median hard-coral cover of 43.1% nearshore and 44.0% offshore. Higher turbidity/productivity was associated with fewer fish and invertebrates but higher invertebrate diversity. Hard-coral cover was positively associated with fish counts and grouper/parrotfish biomass. MPA status did not predict higher diversity, abundance or this biomass in the sampled reefs. Read the results and their qualifications.

These are comparisons among surveyed reefs. One visit per reef does not establish recovery, total local populations, or a causal protection effect. Identification categories include broader groups and pooled “other” labels, so category richness is not total species richness.

There are 91 distinct survey identifiers but 90 unique coordinate pairs: Koh Tang_KTA8 and Koh Tang_KTA9 share coordinates and remain separate records. The environmental table has one row per unique coordinate position.

Complete original files

All nine files from the cited dataset are linked here. Six data tables are reproduced in full below; the original downloads preserve their supplied formatting.

• summary_metadata.csv (provider copy) — 91 rows, 12 original columns. Survey identifiers, coordinates, dates, depth, visibility and management categories.
• benthic_multivariate.csv (provider copy) — 91 rows, 25 original columns. All 16 benthic cover-category columns.
• dbrda_r2.R (provider copy). Original community-composition analysis code.
• fish_multivariate.csv (provider copy) — 91 rows, 69 original columns. All 62 fish identification-group count columns.
• envvar.csv (provider copy) — 90 rows, 18 original columns. Environmental summaries for the 90 unique coordinate positions.
• maindf.csv (provider copy) — 91 rows, 15 original columns. Hard-coral cover, diversity, observed counts and grouper/parrotfish biomass.
• inverts_multivariate.csv (provider copy) — 91 rows, 36 original columns. All 29 invertebrate identification-group count columns, including zero-only columns.
• glmm_r2.R (provider copy). Original diversity, abundance and biomass model code.
• README.md (provider copy). Original deposited file descriptions and variable notes.

Download the complete deposited dataset and code

Reading the original fields

Survey identifiers, dates and coordinates

siteID joins the survey tables. Location names one of the 11 location groups; Survey.site retains the survey label. Lat and Long are decimal degrees. Dates use day/month/year. Depth_m and Visibility_m are metres.

Environmental and management fields

PC1 and PC2 are dimensionless environmental component scores. Category retains On (nearshore) and Off (offshore). Protection is the source Yes/No designation. MPA Zoning or MPA.Zoning records the original management category.

Cover, diversity and observed counts

HC is hard-coral cover as a proportion from 0 to 1; multiply by 100 for percent. fish_simpson and invert_simpson are Simpson diversity (1−D), from 0 to 1. fish_sp_abun and invert_sp_abun count occupied identification groups. fish_total_abun and invert_total_abun are individuals counted. total_site_biomass is estimated grouper and parrotfish biomass in grams, not biomass of all fish.

Fish and invertebrate category columns

Each named category contains individuals counted during the survey. Fish has 62 count columns; invertebrates has 29, including the retained zero-only Squid, Lobster and Chocolate.Chip.Star columns. Original labels, including sum_ fields, are preserved. They are counts, not biomass.

Benthic category columns

The 16 category fields are cover proportions from 0 to 1: CE, CF, CMR, CV, SC, SP, DC, BR, RB, SD, SI, TA, MA, OT, branching and massive. Source codes remain unchanged; the paper describes its benthic classification in the survey methods.

Satellite environmental summaries

kd490_* is the attenuation coefficient at 490 nm, in m⁻¹; sst_* is sea surface temperature in °C; chla_* is chlorophyll a in mg m⁻³. Suffixes max, min, mean and sd denote maximum, minimum, mean and standard deviation.

The original README swaps the benthic and invertebrate file dimensions; the tables here follow the actual deposited files. It also gives 2004–2014 for the satellite summaries, whereas the paper specifies January 2014–January 2024. That period disagreement remains unresolved. These summaries are long-term environmental context, separate from the 2021–2023 reef visits.

Blank entries and NA retain the original missing-value notation. Recorded zeros remain zero. Read the original variable notes · Provider copy.

All original data rows

Open a table to read every deposited row and field. Wide tables scroll horizontally. Values and row order are preserved; no coincident survey or zero category has been removed.