In cdmuir/bayCi: Bayesian Photosynthetic Response Curves Using Stan

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

library(bayCi)
library(dplyr)
library(gunit)
library(units)

In this vignette we will analyze a single Rapid $A-C_i$ Response (RACiR) curve using the braycir() (pronounced like "bracer"). The example data associted with the bayCI package comes from Stinziano et al. (2017). It contains RACiR curves from Poplar deltoides and the object is called stinziano_etal_2017. It can accessed once bayCI is loaded into your environment.

Step 1: Convert data to the racir class

The braycir() function is for fitting a single RACiR curve. The data argument in the braycir() function must inherit class racir. This ensures that all required variables:

• are present in the data
• have the correct column name
• have proper scientific units

The racir() function will make a new racir object from a data.frame and validate it.

df <- stinziano_etal_2017 %>%

  # Set units
  mutate(
    A     = set_units( A,         umol / m^2 / s ),
    Tleaf = set_units( Tleaf,     degreeC        ),
    Q     = set_units( Q,         umol / m^2 / s ),
    E     = set_units( E,         mol / m^2 / s  ),
    gsc   = set_units( gsw / 1.6, mol / m^2 / s  ),
    gtc   = set_units( gtc,       mol / m^2 / s  ),
    CO2_r = set_units( CO2_r,     umol / mol     ),
    CO2_s = set_units( CO2_s,     umol / mol     ),
    Pa    = set_units( Pa,        kPa            ),
    Pci   = set_units( Pci,       Pa             )
  ) %>%

  # Following Sharkey et al. (2007), bayCi uses Pa units
  mutate(
    Pcr = set_units(CO2_r * Pa, Pa),
    Pcs = set_units(CO2_s * Pa, Pa),
    Pgsc = convert_conductance(gsc, P = Pa, Temp = Tleaf)$`umol/m^2/s/Pa`
  ) %>%

  select(A, Cr = CO2_r, Cs = CO2_s, E, gsc, gtc, Pa, Pcr) %>%

  filter(Cr > set_units(10, umol / mol), Cr < set_units(480, umol / mol))

rcr <- racir(df)

class(rcr)

rcr %>%
  select(A, Pcr) %>%
  mutate_if(is.numeric, signif, digits = 3) %>%
  head() %>%
  knitr::kable()

Step 2: Convert empty chamber RACiR correction data to the empty class

Raw RACiR data be corrected using a "blank" curve measured on an empty chamber. These data are passed to braycir() using the empty, which must inherit class empty. This performs the same set of checks as the racir class. The empty() function will make a new empty object from a data.frame and validate it. The correction curve from Stinziano et al. (2017) comes with the bayCi package and can be accessed from stinziano_etal_2017_empty.

df <- stinziano_etal_2017_empty %>%

  # Set units
  mutate(
    A    = set_units( A,    umol / m^2 / s ),
    Cr   = set_units( Cr,   umol / mol     ),
    time = set_units( time, s              )
  )

mty <- empty(df)

class(mty)

mty %>%
  mutate_if(is.numeric, signif, digits = 3) %>%
  head() %>%
  knitr::kable()

References

Sharkey TD, CJ Bernacchi, GD Farquhar, EL Singsaas. 2007. Fitting photosynthetic carbon dioxide response curves for C$_3$ leaves. Plant, Cell & Environment 30(9): 1035--1040.

Stinziano JR, PB Morgan, DJ Lynch, AJ Saathoff, DK McDermitt, DT Hanson. 2017. The rapid $A-C_i$ response: photosynthesis in the phenomic era. Plant, Cell & Environment 40(8): 1256--1262.

cdmuir/bayCi documentation built on Jan. 19, 2020, 12:27 a.m.