Skip to content
/ CSP Public

Data and R code accompanying article 10.1111/gcb.16299 in Global Change Biology

License

Notifications You must be signed in to change notification settings

lukaseamus/CSP

Repository files navigation

Climate-driven shifts in kelp forest composition reduce carbon sequestration potential

This repository contains data and annotated R code accompanying article 10.1111/gcb.16299 in Global Change Biology, split into five folders. Assimilation, Export, Decomposition and Sequestration contain all files on carbon assimilation, export, remineralisation and potential sequestration. Irradiance contains files required to model the local light regime by depth, a prerequisite for cumulative detrital carbon assimilation estimation in Sequestration. Below is a description of each file within those folders as well as the input and output of each R script. The published Manuscript.pdf, Supplement.pdf and Figures in vector format are also provided.

Assimilation

  1. Assimilation.csv: Net and gross carbon assimilation data.
    • species = categorical variable with levels Laminaria digitata (d), Laminaria hyperborea (h) and Laminaria ochroleuca (o)
    • bag = random factor (categorical variable) with plants (P3...24) and mesh bags (B1...9) as levels
    • age = detrital age given in days
    • R = respiration rate given in µmol oxygen per gram of buoyant mass per hour (buoyant mass is practically identical to wet mass)
    • NPP = net photosynthesis rate given in µmol oxygen per gram of buoyant mass per hour
    • GPP = gross photosynthesis rate given in µmol oxygen per gram of buoyant mass per hour (NPP + R)
    • d:w = dry to wet mass ratio
  2. Assimilation.R: Code to analyse and visualise carbon assimilation.
    • Input = Assimilation.csv, Figure 5b from Decomposition.R
    • Output = Figure 5, Figure S8, carbon assimilation functions and results

Export

  1. Export.csv: Carbon export data.
    • month = month and year given as MMM-YY
    • season = categorical variable with levels Spring, Summer, Autumn and Winter
    • time = numerical expression of months
    • species = categorical variable with levels Laminaria digitata (d), Laminaria hyperborea (h) and Laminaria ochroleuca (o)
    • dw.export = biomass export given in grams of dry mass per plant per day
    • fw.export = biomass export given in grams of wet mass per plant per day, converted from dry mass with plant-specific dry to wet mass ratios
    • fw.export.avg = biomass export given in grams of wet mass per plant per day, converted from dry mass with species- and month-specific dry to wet mass ratios
    • C.export = carbon export given in grams per plant per day, converted from dry mass with species- and month-specific carbon content (%)
  2. Carbon.csv: Lamina carbon content data.
    • month = month and year given as MMM-YY
    • season = categorical variable with levels Spring, Summer, Autumn and Winter
    • species = categorical variable with levels Laminaria digitata (d), Laminaria hyperborea (h) and Laminaria ochroleuca (o)
    • carbon = carbon content (%)
  3. Mass.csv: Sporophyte mass data.
    • month = month and year given as MMM-YY
    • season = categorical variable with levels Spring, Summer, Autumn and Winter
    • species = categorical variable with levels Laminaria digitata (d), Laminaria hyperborea (h) and Laminaria ochroleuca (o)
    • mass = whole plant wet mass given in grams
  4. DW.csv: Dry to wet mass ratio data.
    • month = month and year given as MMM-YY
    • season = categorical variable with levels Spring, Summer, Autumn and Winter
    • species = categorical variable with levels Laminaria digitata (d), Laminaria hyperborea (h) and Laminaria ochroleuca (o)
    • d.w = dry to wet mass ratio
  5. Density.csv: Sporophyte density data.
    • month = month and year given as MMM-YY
    • season = categorical variable with levels Spring, Summer, Autumn and Winter
    • species = categorical variable with levels Laminaria digitata (d), Laminaria hyperborea (h) and Laminaria ochroleuca (o)
    • density = number of plants per square metre. Note that Laminaria hyperborea and Laminaria ochroleuca share each quadrat while Laminaria digitata occurs spatially separated.
  6. Export.R: Code to analyse and visualise carbon export.
    • Input = Export.csv, Carbon.csv, Mass.csv, DW.csv, Density.csv
    • Output = Figure S2, Constants.csv, carbon export results

Decomposition

  1. Decomposition.csv: Biomass decomposition data.
    • species = categorical variable with levels Laminaria digitata (d), Laminaria hyperborea (h) and Laminaria ochroleuca (o)
    • site = categorical variable with levels West Hoe (WH), Drake's Island (DI) and Jennycliff (JC)
    • substratum = categorical variable with levels Forest and Sediment
    • mesh = mesh diameter given in centimetres
    • g.loss = absolute biomass loss given in grams per day
    • perc.loss = relative biomass loss given in percentage of initial mass per day
  2. Biochemical.csv: Elemental stoichiometry and phenols in relation to decomposition.
    • species = categorical variable with levels Laminaria digitata (d), Laminaria hyperborea (h) and Laminaria ochroleuca (o)
    • bag = random factor (categorical variable) with mesh bags (B1...9) as levels
    • age = detrital age given in days
    • g.loss = absolute biomass loss given in grams per day
    • perc.loss = relative biomass loss given in percentage of initial mass per day
    • phenols = final soluble polyphenolic content (%)
    • N = final nitrogen content (%)
    • C = final carbon content (%)
    • CN = final carbon to nitrogen ratio
  3. Grazing.csv: Image analysis data of tissue damage on final retrieval.
    • species = categorical variable with levels Laminaria digitata (d), Laminaria hyperborea (h) and Laminaria ochroleuca (o)
    • bag = random factor (categorical variable) with mesh bags (B1...9) as levels
    • excavation = surface area of excavation scars relative to total tissue area (%)
    • perforation = surface area of holes relative to total tissue area plus holes (%)
  4. Decomposition.R: Code to analyse and visualise carbon export.
    • Input = Decomposition.csv, Biochemical.csv, Grazing.csv
    • Output = Figure 3, Figure S3, decomposition results

Irradiance

  1. L4.csv: Physical and chemical data from station L4, compiled from data available at https://www.westernchannelobservatory.org.uk/l4_ctdf/index.php.
    • Date = date given as DD.M.YY
    • Month = month
    • Year = year given as YYYY
    • Season = categorical variable with levels Spring, Summer, Autumn and Winter
    • Temp = temperature (°C)
    • Fluor = fluorescence given in milligrams of chlorophyll a per cubic metre
    • Depth = depth given in metres
    • Density = water density given in kilograms per cubic metre
    • Salinity = salinity (‰)
    • Trans = transmission (%)
    • PAR = photosynthetically active radiation given in µmol photons per square metre per second
    • Oxygen = oxygen given in µM
    • Sound = sound velocity given in metres per second
  2. Irradiance.R: Code to analyse the depth-irradiance relationship.
    • Input = L4.csv
    • Output = seasonal and annual exponential depth-irradiance relationships

Sequestration

  1. Environmental.csv: Physical data from the West Hoe decomposition experiment.
    • date = date given as DD/MM/YYYY
    • time = time given as HH:MM:SS
    • d = days from start of expeeriment given as integers
    • day = days expressed numerically
    • sunrise = time of sunrise given as HH:MM:SS
    • sunset = time of sunset given as HH:MM:SS
    • daytime = categorical variable with levels day and night based on sunrise and sunset times
    • temp = temperature (°C)
    • lux = light intensity given in lux
  2. Constants.csv: Species-specific export and density data compiled in Export.R.
    • period = categorical variable with levels Year, Spring, Summer, Autumn and Winter
    • species = categorical variable with levels Laminaria digitata (d), Laminaria hyperborea (h) and Laminaria ochroleuca (o)
    • biomass = biomass export given in grams of dry mass per plant per year
    • carbon = carbon export given in grams per plant per year
    • density = number of plants per square metre
    • bCI = half 95% confidence interval (z × standard error) of the product of biomass export and density
    • cCI = half 95% confidence interval (z × standard error) of the product of carbon export and density
  3. Sequestration.R: Code to analyse the depth-irradiance relationship.
    • Input = Environmental.csv, Constants.csv, gross carbon assimilation functions from Assimilation.R, photosynthesis-irradiance relationship for Laminaria hyperborea (doi 10.3354/ab00515), sunlight lux to µmol photons per square metre per second conversion, seasonal and annual exponential depth-irradiance relationships from Irradiance.R, seasonal and annual daylight hours for Plymouth, local carbon sink coordinates (doi 10.1002/ecm.1366), HadISST historical sea surface tempertaure data available at https://www.metoffice.gov.uk/hadobs/hadisst/data/download.html (doi 10.1029/2002jd002670), Bio-ORACLE representative concentration pathway sea surface temperature predictions available at https://www.bio-oracle.org/downloads-to-email.php (doi 10.1111/geb.12693), species-specific thermal tolerance data from Figure 2b and Table S4
    • Output = Figure 4, Figure S4, Figure S5, Figure S6, Figure S7, carbon export estimates, carbon sequestration potential, cumulative detrital carbon assimilation

Luka Seamus Wright, 10 July 2022