Published February 24, 2021 | Version v.1.0
Software Open

fionasoper/Nitrogen-fixation-methods: Supplemental R Code for Soper et al. 2021

Creators

Description

Accompanying R code for simulating and analyzing biological nitrogen fixation data, for 'Soper et al. 2021. A roadmap for sampling and scaling biological nitrogen fixation in territorial ecosystems. Methods in Ecology and Evolution.'

Code by Benton Taylor, Harvard University.

 

Part 1. Simulating and scaling symbiotic BNF data generated by 15N2 incorporation assays and field nodule biomass sampling

This code simulates data frames (of nodule biomass and BNF rate per unit nodule biomass) that can be used 1) for the user to apply test statistics, and 2) as sample data for subsequent scaling steps described below. The data presented in the simulation is representative of sampling conducted in lowland tropical rainforests. However, users can change any of these input values to represent their own sampling site.

  1. Data frame simulation for nodule biomass from four plots, A-D, approximating  a stratified adaptive sampling cluster scheme where additional cores are taken in the vicinity of cores that are found to contain nodules. Inputs are number of cores sampled per plot and proportion of cores containing nodules. The data generated is nodule biomass in kg dry mass ha-1, following an approximately zero-inflated log-normal distribution.
  2. Data frame simulation for generating results of 15N2 incubations on a subset of cores (10 per plot, all nodules per core pooled) from the simulated data frame above. Generates data as d15N values.
  3. Code to convert data from 1.2 from d15N values to mass specific BNF rates (g N g nodule dry mass-1 yr-1). Inputs are incubation duration (hours), dry mass of nodules (g), proportion of chamber headspace replaced with 15N2 and atom percent of labelled 15N2 incubation gas.
  4. Code to convert symbiotic 15N incorporation assay data to plot-scale BNF. This code uses average mass-specific BNF rates from 1.3 and nodule biomass data from 1.1 to scale rates to kg N ha-1 y-1. Additional data inputs are core diameter (cm). Error from each term is propagated as 1 standard deviation.

Part 2. Simulating and scaling symbiotic BNF data generated by the 15N2 dilution technique

This code simulates a data frame of BNF rates derived from the 15N2 dilution technique that can be used 1) for the user to apply test statistics, and 2) as sample data for %Ndfa calculations described below.

  1. Data frame simulation for 15N2 dilution scenario comparing foliar d15N for a fixing and non-fixing reference plants. Data inputs are foliar d15N and %N values for fixing and reference species.
  2. Code for calculating %nitrogen derived from fixation (%Ndfa) using a two end-member mixing model from paired fixing/non-fixing plant pairs using data generated in 2.1, under three scenarios
    1. Input is a single fixed value for the d15N of N derived from fixation, and a single foliar d15N value for one reference plant.
    2. Input is a single fixed value for the d15N of N derived from fixation, and a distribution (mean and standard deviation) of foliar d15N values for multiple reference plants.
    3. Input is a distribution (mean and standard deviation) of values for the d15N of N derived from fixation, and a distribution (mean and standard deviation) of foliar d15N values for multiple reference plants.

 

Files

fionasoper/Nitrogen-fixation-methods-v.1.0.zip

Files (5.5 kB)

Name Size Download all
md5:e9ce75a8fcf39718ca7bb6bcbe4e4d36
5.5 kB Preview Download

Additional details