Intermittent-flow respirometry: Simple example • respR

Intermittent-flow respirometry

Intermittent-flow respirometry involves obtaining multiple replicates in an experiment by periodically flushing the respirometer with fresh treatment water or air after a set period of time or O₂ decrease. Typically, the oxygen sensing system is left recording during flushes so the system returning to normoxia can be seen in the data. If the dataset is not split into the separate replicates, this results in an dataset with oscillating levels of oxygen. This vignette shows a simple example of how an intermittent-flow respirometry experiment can be analysed in respR. Another example is provided to show how a longer, more complex intermittent-flow experiment with differing background rates can be analysed.

A relatively simple example

The example data, intermittent.rd, contains intermittent-flow measurements of oxygen consumption in the sea urchin Heliocidaris erythrogramma. Detailed information about the data, including its source and methods, can be obtained with the command ?intermittent.rd.

We use inspect() to scan for errors in the data frame, preview the data and save it to an object to pass to the next stage.

urchin_int <- inspect(intermittent.rd)

#> inspect: Applying column default of 'time = 1'
#> inspect: Applying column default of 'oxygen = 2'
#> inspect: No issues detected while inspecting data frame.

#> 
#> # print.inspect # -----------------------
#>               Time   O2
#> NA/NAN        pass pass
#> sequential    pass    -
#> duplicated    pass    -
#> evenly-spaced pass    -
#> 
#> -----------------------------------------

We can see from the top plot that this dataset has three replicates. The flush periods obviously skew the bottom plot, which shows rate of change (positive or negative) in O₂ concentration over a rolling window of 10% of the data. However specimen uptake rates (flat regions) appear to be relatively consistent within each experimental replicate.

Analysis with calc_rate()

In the near future respR will support intermittent-flow data directly with a dedicated function to easily identify and subset replicates, but at present intermittent data can be analysed quite easily via the main calc_rate() function with a little organisation.

Subsetting by row and time

The calc_rate function supports input of multiple subset regions by row or time. The from and to operators can be vectors with paired values:

urchin_int_rates <- calc_rate(urchin_int, from = c(200, 2300, 4100), to = c(1800,
    3200, 4600), by = "time")
#> 
#> # plot.calc_rate # ----------------------
#> calc_rate: Plotting rate from position 1 of 3 ... 
#> To plot others use 'pos'

#> -----------------------------------------
print(urchin_int_rates)
#> 
#> # print.calc_rate # ---------------------
#> Rank 1 of 3 rates:
#> Rate: -0.0005734109 
#> 
#> To see other results use 'pos' input. 
#> To see full results use summary().
#> -----------------------------------------

Calling summary() will show the coefficients and locations of each subset:

summary(urchin_int_rates)
#> 
#> # summary.calc_rate # -------------------
#> Summary of all rate results:
#> 
#>    rank intercept_b0       rate_b1   rsq  row endrow time endtime  oxy endoxy
#> 1:    1     7.127202 -0.0005734109 0.995  201   1801  200    1800 7.05   6.11
#> 2:    2     8.528274 -0.0006097325 0.990 2301   3201 2300    3200 7.12   6.58
#> 3:    3     9.731918 -0.0006539752 0.967 4101   4601 4100    4600 7.08   6.73
#>      rate.2pt          rate
#> 1: -0.0005875 -0.0005734109
#> 2: -0.0006000 -0.0006097325
#> 3: -0.0007000 -0.0006539752
#> -----------------------------------------

We can also use plot() to view the diagnostic plots of any individual replicate. The argument rep determines which data subset to plot:

plot(urchin_int_rates, rep = 2)
#> 
#> # plot.calc_rate # ----------------------
#> calc_rate: Plotting rate from position 1 of 3 ... 
#> To plot others use 'pos'

#> -----------------------------------------

We can even query multiple regions of each replicate (which can overlap, although this would obscure the output plot somewhat):

calc_rate(urchin_int, from = c(200, 900, 2300, 2900, 4000, 4400), to = c(700, 1700,
    2700, 3300, 4300, 4700), by = "time")
#> 
#> # plot.calc_rate # ----------------------
#> calc_rate: Plotting rate from position 1 of 6 ... 
#> To plot others use 'pos'

#> -----------------------------------------
#> 
#> # print.calc_rate # ---------------------
#> Rank 1 of 6 rates:
#> Rate: -0.0006790314 
#> 
#> To see other results use 'pos' input. 
#> To see full results use summary().
#> -----------------------------------------

Extracting rate results

Rates can be easily extracted for further analysis or selection. For instance, we may be interested in the lowest rate, the one most representative of basal metabolic rates. Of course because rates are negative, the minimum value in negative terms would actually represent the highest rate, so we should extract the maximum (least negative) value, or transform the rates first:

urchin_int_rates$rate
#> [1] -0.0005734109 -0.0006097325 -0.0006539752
min(-1 * urchin_int_rates$rate)
#> [1] 0.0005734109

Or, we may be more interested in the mean rate:

mean(urchin_int_rates$rate)
#> [1] -0.0006123729

Subsetting by oxygen

If we are interested in looking at rates over particular oxygen ranges within each replicate, calc_rate() will find the first match to the range entered:

calc_rate(urchin_int, from = 7, to = 6.6, by = "o2")
#> 
#> # plot.calc_rate # ----------------------
#> calc_rate: Plotting rate from position 1 of 1 ...

#> -----------------------------------------
#> 
#> # print.calc_rate # ---------------------
#> Rank 1 of 1 rates:
#> Rate: 6.404498e-05 
#> 
#> To see full results use summary().
#> -----------------------------------------

To examine the same range in each replicate, we can follow a number of approaches. We can call the data range containing another replicate directly from the original dataset, or data frame saved within the inspect() object:

calc_rate(intermittent.rd[2100:3500, ], from = 7, to = 6.6, by = "o2")
#> 
#> # plot.calc_rate # ----------------------
#> calc_rate: Plotting rate from position 1 of 1 ...
#> -----------------------------------------
#> 
#> # print.calc_rate # ---------------------
#> Rank 1 of 1 rates:
#> Rate: -0.0006099152 
#> 
#> To see full results use summary().
#> -----------------------------------------
calc_rate(urchin_int$dataframe[2100:3500, ], from = 7, to = 6.6, by = "o2")
#> 
#> # plot.calc_rate # ----------------------
#> calc_rate: Plotting rate from position 1 of 1 ...
#> -----------------------------------------
#> 
#> # print.calc_rate # ---------------------
#> Rank 1 of 1 rates:
#> Rate: -0.0006099152 
#> 
#> To see full results use summary().
#> -----------------------------------------

Or we can subset them manually into new dataframes, and process them separately through the entire workflow. Here we use the dedicated subset_data() function in respR, which allows easy subsetting by time, oxygen or row number. For best results, these should exclude the flush periods.

# Create separate replicate data frames
u_rep1 <- subset_data(intermittent.rd, from = 1, to = 1900, by = "time")
u_rep2 <- subset_data(intermittent.rd, from = 2100, to = 3500, by = "time")
u_rep3 <- subset_data(intermittent.rd, from = 3700, to = 4831, by = "time")

# Process one in `inspect()` (output not shown)
u_rep3_c <- inspect(u_rep3)

calc_rate(u_rep3_c, from = 7, to = 6.6, by = "o2")
#> 
#> # plot.calc_rate # ----------------------
#> calc_rate: Plotting rate from position 1 of 1 ...

#> -----------------------------------------
#> 
#> # print.calc_rate # ---------------------
#> Rank 1 of 1 rates:
#> Rate: -0.0006142321 
#> 
#> To see full results use summary().
#> -----------------------------------------

Analysis with auto_rate()

To more rigourously identify minimum, maximum or most linear rates we can use the auto_rate() function, however because this does not support (as yet) subsetting internally, each replicate must be subset and analysed separately. See other vignettes for more detail on using auto_rate().

## subset first replicate
u_rep1 <- subset_data(intermittent.rd, from = 0, to = 1900, by = "time")

## analyse in auto_rate
auto_rate(u_rep1)  #default = most linear rate

auto_rate(u_rep1, method = "min")  # minimum rate
#> Warning in auto_rate(u_rep1, method = "min"): auto_rate: the 'min' and 'max' methods have been deprecated, as they resulted in incorrect ordering of oxygen production rates. 
#>  They have been retained for code compatibility, but may be removed in a future version of respR. 
#>  Please use 'highest/lowest' for ordering by absolute rate value, or 'maximum/minimum' for strict numerical ordering of rates.

Background adjustments

For background adjustments, the adjust_rate() function can be used on calc_rate objects containing multiple rate values:

urchin_int_rates_adj <- adjust_rate(urchin_int_rates, by = -5e-05)
#> adjust_rate: Rate adjustments applied using "mean" method. 
#> Use print() or summary() on output for more info.
urchin_int_rates_adj$input.rate
#> NULL
urchin_int_rates_adj$corrected
#> NULL

We can see each individual rate has been adjusted, and the new rates can be extracted from the resulting object using $corrected.

Converting rates

Once the rates have been determined, they can be converted to volume-corrected or mass-specific rates using convert_rate().

u_rates_final <- convert_rate(urchin_int_rates_adj, 
                              o2.unit = "mg/L", 
                              time.unit = "s", 
                              output.unit = "mg/h/g", 
                              volume = 2.379, 
                              mass = 0.006955)
#> convert_rate: object of class `adjust_rate` detected. Converting all adjusted rates in '$rate.adjusted'.

Only the first rate converted rate is shown in the output, but all rates determined are converted and can be extracted.

u_rates_final$input
#> $x
#> 
#> # print.adjust_rate # -------------------
#> NOTE: Consider the sign of the adjustment value when adjusting the rate.
#> 
#> Adjustment was applied using the 'mean' method.
#> 
#> Rank 1 of 3 adjusted rate(s):
#> Rate          : -0.0005734109
#> Adjustment    : -5e-05
#> Adjusted Rate : -0.0005234109 
#> 
#> To see other results use 'pos' input.
#> To see full results use summary().
#> -----------------------------------------
#> 
#> $o2.unit
#> [1] "mg/L"
#> 
#> $time.unit
#> [1] "sec"
#> 
#> $output.unit
#> [1] "mgO2/hour/g"
#> 
#> $volume
#> [1] 2.379
#> 
#> $mass
#> [1] 0.006955
#> 
#> $area
#> NULL
#> 
#> $S
#> NULL
#> 
#> $t
#> NULL
#> 
#> $P
#> [1] 1.013253
u_rates_final$output
#> [1] "mgO2/hour/g"

More complex intermittent-flow experiments

A more complex intermittent-flow analysis is provided here.