Dataset DOI: 10.5061/dryad.sqv9s4njd

Description of the data and file structure

The data were collected as part of a field study evaluating ecological responses to kelp forest restoration in the Gwaii Haanas archipelago of Haida Gwaii, British Columbia, Canada. The study used a Before–After–Control–Impact (BACI) design to compare a kelp restoration site, in which sea urchins were manually reduced, with a nearby unrestored control site. Restoration activities occurred in 2018 and 2019 and were followed by annual monitoring surveys conducted from 2017 through 2021. Divers carried out subtidal surveys along permanent transects spanning multiple depth zones to quantify benthic community structure, including densities and size structure of macroalgae, invertebrates, and fish. Tissue samples from key herbivores, the red sea urchin (Mesocentrotus franciscanus) and the northern abalone (Haliotis kamtschatkana), were collected and analyzed for fatty acid composition using gas chromatography mass spectrometry to characterize trophic linkages between recovering kelp forests and consumer diets. Together, these data document changes in kelp abundance, grazer population metrics, and fatty acid biomarkers before and after restoration, allowing evaluation of how kelp recovery alters the nutritional landscape and physiological condition of key consumers.

Note: The uploaded files contain no latitude or longitude data, and site names are intentionally generalized to obscure precise locations. Instead, locations are described only at coarse scales (e.g., Haida Gwaii / Gwaii Haanas, control versus restoration sites, and broad depth ranges), and sampling was conducted near—but not directly on—permanent transects to avoid identifying exact specimen positions. The study area lies within the protected Gwaii Haanas National Park Reserve, National Marine Conservation Area Reserve, and Haida Heritage Site, which is co-managed by the Council of the Haida Nation and the Government of Canada. Fisheries and Oceans Canada (DFO), particularly its Conservation and Protection Branch, enforces fishery regulations through patrols, inspections, and investigations, while Parks Canada holds additional responsibility under the Species at Risk Act. Because illegal harvest is the primary threat to northern abalone, detailed geospatial information is intentionally withheld for conservation purposes.

Files and variables

Missing values

Missing values are recorded as NA. Blank cells may occur in some raw survey files when a variable was not recorded or was not applicable.

Units are reported in variable names where applicable (e.g., .mm, .cm, .m, .m2, .g). Density values are reported as individuals per square meter unless otherwise specified.

Fatty acid variables represent proportional abundance relative to total fatty acids in each sample.

---

Benthic monitoring surveys

• 2017_09_Sep_Fish_Permanent_20191009.csv
• 2017_09_Sep_Macroinvert_Permanent_20191011.csv
• 2017_09_Sep_Quadrats_Permanent_20191106.csv
• 2017_09_Sep_UrchAb_Permanent_20191008.csv
• 2018_07_Jul_Fish_Permanent_20191009.csv
• 2018_07_Jul_Macroinvert_Permanent_20191011.csv
• 2018_07_Jul_Quadrats_Permanent_20190907.csv
• 2018_07_Jul_UrchAb_Permanent_20191008.csv
• 2018_09_Sep_Fish_Permanent_20191009.csv
• 2018_09_Sep_UrchAb_Permanent_20191008.csv
• 2018_10_Oct_UrchAb_Permanent_20191008.csv
• 2019_03_Mar_UrchAb_Permanent_20191008.csv
• 2019_07_Jul_Fish_Permanent_20191009.csv
• 2019_07_Jul_Macroinvert_Permanent_20191011.csv
• 2019_07_Jul_Quadrats_Permanent_20190915.csv
• 2019_07_Jul_UrchAb_Permanent_20191008.csv
• 2020_Macroinvert_Permanent_20210316_CHECKED.csv
• 2020_UrchAb_Permanent_20210315_CHECKED.csv
• 2021_Macroinvert_Permanent_20211020_Final.csv
• 2021_UrchAb_Permanent_20211020_Final.csv

Standard benthic monitoring survey variables

Many survey datasets share the following variables:

Year, Month, Day
Date of survey.

Site
Study site identifier.

Plot
Plot nested within site.

Treatment
Site designation. Restoration or Control.

Pre.Post / Period
Indicator of sampling relative to restoration intervention.

TransectLevel
Depth stratum of transect (shallow, mid, deep).

TransectLength.m, TransectWidth.m
Transect dimensions in meters.

QuadArea.m2 or BandArea.m2
Area sampled in square meters.

Species
Taxonomic identification.

Size.mm, Size.cm
Individual body size.

Count
Number of individuals observed.

Density
Calculated density per square meter.

FuncGroup
Functional group classification.

---

Fatty acid variables

Variables beginning with C represent fatty acids identified by carbon chain length and unsaturation (for example, C18.1w9c). Values represent proportional abundance in each sample.

Examples include:

C16.0
Palmitic acid

C18.1w9c
Oleic acid

C20.5w3
Eicosapentaenoic acid

Variables beginning with i. or a. represent iso and anteiso branched fatty acids.
Variable other represents the remaining unresolved fatty acid fraction.

File: Quad_Algae.csv

Description
Quadrat-level survey data for macroalgae collected along permanent transects at restoration and control sites. Data include species identity, size measurements, abundance, and derived density estimates.

Additional variables

QuadNum
Quadrat number along the transect.

SampleID
Unique identifier for a biological sample.

Date
Sampling date.

---

File: Quad_Invert.csv

Description
Quadrat-level survey data for small benthic invertebrates recorded along permanent transects.

Variables are consistent with Quad_Algae.csv and include species identity, organism size, abundance, and density estimates.

---

File: Macroinvert.csv

Description
Transect-level observations of macroinvertebrate communities recorded during subtidal surveys.

Additional variables

SourceFile
Original raw data source file.

MonthNum
Numeric month of survey.

TransectID
Unique transect identifier.

TransectUnit
Sampling unit used for density calculation.

YearMonth, YearMonthLab
Combined year-month variables are used for plotting and analysis.

---

File: Kelp.csv

Description
Transect-level survey data for kelp species abundance and size structure.

Variables are consistent with Macroinvert.csv and include transect dimensions, species identity, counts, and density estimates.

---

File: Fish.csv

Description
Fish density observations were recorded along permanent transects.

Additional variables

MonthNum
Numeric month of survey.

YearMonth
Combined year and month variable.

SurveyType
Fish survey protocol.

TransectUnit
Sampling unit used for density calculations.

---

File: Urchin_Gonad_Mass_Data.csv

Description
Morphological and reproductive measurements of red sea urchins were used to estimate body condition.

Variables

test_diam_mm
Urchin test diameter in millimeters.

gonad_mass_g
Gonad mass in grams.

site
Study site.

date
Sampling date.

depth
Depth category or transect level.

treatment
Restoration or control site.

year
Year of sampling.

---

File: HAKA_Fatty_Acid_Data.csv

Description
Fatty acid composition of consumer tissues collected from northern abalone and red sea urchins.

Additional variables

year
Year sample collected.

island
Island or site location.

habitat
Habitat classification.

long_axis_mm
Shell or test long axis length.

short_axis_mm
Shell short axis length.

shell_ht_mm
Shell height.

wet_mass_g
Wet mass of an organism.

gonad_color
Observed gonad color.

sex
Sex of the organism is identifiable.

sample_id
Unique sample identifier.

Fatty acid variables are described above.

---

File: MEFR_Fatty_Acid_Data.csv

Description
Fatty acid composition of red sea urchin tissues across habitats and treatments.

Additional variables

year
Year sample collected.

island
Island or site location.

habitat
Habitat classification.

long_axis_mm
Urchin test diameter.

treatment
Restoration or control designation.

island_habitat
Combined location and habitat variable.

Fatty acid variables follow the format described above.

---

File: Algae_Fatty_Acid_Data.csv

Description
Fatty acid composition of macroalgae collected to characterize biochemical signatures of primary producers.

Additional variables

year_collected
Year algae sample collected.

date_collected
Collection date.

fa_extraction_date
Date of fatty acid extraction.

dna_id
Sample identifier.

gcms_batch_id
Gas chromatography mass spectrometry batch identifier.

sample_id
Unique sample identifier.

lat, long
Latitude and longitude of the collection site.

site
Sampling location.

depth_ft
Collection depth in feet.

exp_time_point
Experimental time point.

habitat
Habitat classification.

division, class, order, family, genus, species
Taxonomic classification of algae.

tissue_type
Tissue analyzed.

Fatty acid variables follow the format described above.

---

Raw survey files

Files named with the format

YYYY_MM_[SurveyType]Permanent[date].csv

contains the original diver survey records used to generate the processed datasets above.

These files include raw observations of fish, macroinvertebrates, algae, sea urchins, and abalone collected along permanent transects.

Common variables include:

Year, Month, Day
Date of survey.

Observer
Survey diver.

SurveyType
Survey protocol.

TransectLevel
Depth stratum.

TransectLength.m, TransectWidth.m
Transect dimensions.

TimeStart, TimeEnd
Survey start and end time.

SurveyMinutes
Survey duration.

Depth.ft or Depth.m
Depth of survey.

Substrate variables
Dominant substrate categories.

KelpCover and KelpCoverSp variables
Visual estimates of kelp cover.

Species
Species identification.

Size.mm or Size.cm
Individual size.

Count
Number of individuals observed.

Notes
Free text field describing observations or anomalies.

Additional variables such as Behaviour, Cover, TagNumber, Sex, SizeEstimated, and PointCount appear in specific survey types.

Code/software

Code available from Zenodo (https://doi.org/10.5281/zenodo.18407196).

All analyses were conducted using the R statistical computing environment (R version 4.x) within RStudio using the included project file (Biomarkers_of_Recovery.Rproj). The repository contains R scripts used for data processing, statistical analyses, and figure generation associated with the manuscript.

To reproduce the analysis workflow, users should install R and the CRAN packages listed in the script CRAN_Libraries.R, which loads the required libraries. Major packages used include:

readxl
Hmisc
plyr
tidyverse
janitor
forcats
stringr
reshape2
tidybayes
ggrepel
cowplot
egg
RColorBrewer
ggnewscale
rstan
brms
cmdstanr
posterior
boot
vegan
emmeans
car
permute

The general workflow is:

1. Install the required CRAN packages and load them using CRAN_Libraries.R.
2. Run scripts in the Data_Wrangling directory to import and format fatty acid and survey datasets.
3. Run analysis scripts that generate the statistical models and figures used in the manuscript. Key scripts include:
   • Biomarker_Groups.R and Functional_Groups.R for classification of fatty acid biomarkers and ecological functional groups.
   • Algae_Fatty_Acid_Analysis.R, Abalone_Biomarker_Analysis.R, and related scripts for fatty acid analyses.
   • Community_Analyses.R and Kelp_Restoration_Context_Analysis.R for community and restoration analyses.

Raw monitoring survey data (Excel files) are included in the Monitoring_Surveys directory and are imported by the R scripts during preprocessing.

All scripts are plain text R files and can be executed sequentially within the RStudio project environment to reproduce the analytical workflow and figures associated with the dataset.