Respirometry menu

If you are unfamiliar with the fundamentals of respirometry, you might want to read this page before diving into the rest of this section.

These functions convert gas concentrations -- O2, CO2, H2O -- into rates of gas exchange:   VO2, VCO2, evaporative water loss (EWL).   Gas concentrations should be in units of % or optionally for EWL, dew point temperature.   The program expects gas concentration to be expressed as difference from reference levels, with reference (baseline) set to zero (in the PREFERENCES option in the LabHelper menu, you can select automatic lag and baseline correction prior to gas exchange calculations).

Choose one of two computation modes:

• Positive-going deflection:   Changes in gas concentration are positive with respect to baseline (the default).

• Negative-going deflection:  Changes in gas concentration are negative with respect to baseline. This can be used if -- as in oxygen consumption -- gas exchange is measured as a depletion of gas concentration.   Many users find positive-going deflections more intuitive, but the negative-going option is available if desired.

Results are stored either in the source channel or optionally (if there are less than 40 channels) in a new channel.

Gas calculations begin with flow rate source selection.  Flowrate is entered from the file or by the user, or is obtained from a channel (i.e., recorded from a flow meter).

Important:   The program expects flowrates to be in units of ml/min, unless set to liters/min in this window.

For VO2 and VCO2 , you need to specify:

• whether incurrent CO2 is absorbed – note that this refers to gas entering the metabolism chamber, NOT to gas entering the analyzers.   If you are using ‘standard’ air (where the CO2 concentration is about .04%), this option has little effect.

• concentrations of O2 and CO2 in ambient gas (FiO2 and FiCO2; default 20.95% and 0.04%, respectively – note that the program expects these to be entered as percentages, not fractions ).

For EWL calculations (VH2O), and for VO2 and VCO2 if you did not use dry incurrent gas, you will also need to indicate the type of sensor used to measure the water vapor content.   Most lab-grade humidity sensors output relative humidity (the current humidity relative to saturation water content), dew point (the temperature to which air must be cooled to become saturated with water vapor), or water vapor pressure in units of pressure (usually pascals or kilopascal) or mass/volume (usually micrograms/mL, mg/L or grams/m3; all of these are equivalent).

For relative humidity sensors, you will also need to indicate operating temperature of the sensor (this is done later; see below).

Back to top

Channel selection:  In the next window (below,left) you select the channel containing gas concentration data, and if you are obtaining flow from a channel, the flow rate channel:

Flow configurations:  Next you select a flow configuration ('Mode' 1, 2, etc.).  This specifies the conversion equation to be used, based on where flow rate is measured (upstream or downstream of the animal chamber and gas sensor) and how (and if) the the gas stream is dehumidified and scrubbed of CO2.

The 'masks' button (VO2 and VCO2 only) opens a similar window with a selection of mask configurations (example for VO2 shown at right).  A mask system is defined as one in which unscrubbed ambient gas - usually air - is pulled past the animal to capture exhaled gas (i.e., gas is sucked past the animal under negative pressure).  In mask systems all pumps, flow meters, gas analyzers, etc.  are downstream.
       Some mask systems are functionally identical to some 'regular' flow configurations and use the same conversion equations. The '?' button describes these overlaps, and also certain configurations to avoid if possible.

Pick the configuration and mode closest to your own respirometry system.  It's quite important to give LabAnalyst the right information about your flow arrangement.  In some -- but not all -- cases, with 'normal' incurrent concentrations of CO2 and O2 (about 0.04% and 20.9%), different modes yield fairly similar results.  However, in some cases, serious errors (20% or more) can result if the wrong equations are used (this is most problematic for VO2).  Selecting the appropriate configuration is most critical if the sample gas from the animal chamber has large deflections from ambient concentrations of CO2 and O2.  An important and easily avoidable error will result if the gas stream at the flow meter contains a significant (but unknown) fraction of water vapor.  Therefore it's good practice to use DRY gas whenever possible -- and if you can't, be sure to compensate as described in the next section.

Back to top

Compensating for water vapor:  In some flow arrangements, flow rate is measured on a 'wet' gas that is -- usually -- subsequently dried prior to measuring O2 or CO2 concentration (this includes modes 1 and 2 for VO2 and mode 1 for VCO2 [if incurrent gas isn't dry], and most mask systems).  In such cases it may be necessary to compensate for the water vapor content.  It is also necessary to compensate for water content if O2 or CO2 concentration is measured with a 'wet' gas containing water vapor (see the compensating for sensor humidity section).  Compensation is necessary because water vapor affects gas volume.

As can be seen in this diagram, the importance of water vapor compensation is greatest at high ambient temperatures and high humidities.  For example, at a temperature of 30 °C, saturated air (100% R.H.) is about 4 percent water vapor by volume.  That proportion increases to about 7 percent at 40 °C.

Keep in mind that these values are for a total barometric pressure of 1 atmosphere.  Since saturation vapor pressure at a given temperature is constant regardless of the total pressure in the system, the fractional water vapor content increases as total pressure decreases.  For example, if you are working at an altitude of 2000 meters (where atmospheric pressure is roughly 590 torr), saturated air at 30 °C contains about 5.4% water vapor by volume.

You can reduce or avoid these errors by using estimates of humidity, or preferably measurements of humidity or dew point, to calculate the volumetric fraction of incurrent gas comprised of water.  This figure should give you some idea of the need to compensate, given the temperature and pressure conditions you work with.

Note that the temperature and pressures of importance are those at the position of the flowmeter, and not elsewhere in your system.  Clearly, if you are working at fairly low humidities and the temperature of your flowmeter is about 20 °C or lower, the expected error is on the order of 1% or less, and probably can be safely ignored.  Even in fairly warm conditions (around 30 °C), you can reduce the expected error to about 2% if you can guess relative humidity within 50%.

Note that it's usually necessary to adjust to ambient barometric pressure (even if flow rate has already been set to STP).  This is because the partial pressure of water vapor is constant at a given combination of relative humidity and temperature, regardless of the partial pressure of the rest of the gas mixture.   You also have the option of not compensating, if you wish.

You are next asked for a channel containing humidity (or water vapor) data; alternately you can use a constant RH (the default is 20%).  This is shown below, at left.  Next you are next asked whether the humidity data are in units of relative humidity or dew point temperature (below, at right):

   If the data are in units of relative humidity, you are asked to select a channel containing humidity temperature data (as shown at right), or a constant temperature for the RH sensor: 

The software will compensate for temperature differences between the humidity sensor and the flowmeter.

 One other effect of water vapor is (marginally) relevant if you use a mass flow controller or mass flowmeter to handle your flow rates.  These devices ("MF's") determine rates of gas flow by measuring rates of heat transfer.  Unsurprisingly, the thermal properties of water vapor are slightly different from those of dry air, so the addition of water vapor to the air stream will have an effect on the calibration accuracy of MF devices.  However, under most imaginable respirometry setups, this error is small and can safely be ignored.

Back to top

Compensating for RQ:  One of the many troublesome complications in respirometry is that organisms seldom exchange just one gas at a time: instead, they are simultaneously consuming oxygen and producing CO2 (we'll leave water out of this discussion because -- usually -- it can easily be scrubbed from air streams prior to gas analysis).  Since the ratio of CO2 produced / oxygen consumed (the Respiratory Quotient, or RQ) is seldom exactly equal to 1.0, respiration causes a net change in the total gas volume.  As a result, gas concentrations are affected.  For example, consider a calculation of oxygen consumption:   If more oxygen is consumed than CO2 is produced, the total gas volume is slightly reduced, which results in an increase in the concentration of O2.  This seemingly trivial imbalance will influence exchange calculations and needs to be accounted for (startlingly large errors can result if you don't do this correctly).  The necessary equations can be derived from simple algebra, but are tedious in practice (click here for a simple example).  Fortunately, equations for most respirometry systems are built into LabAnalyst (see below for details).

If you are calculating oxygen consumption and you use a flow configuration that requires compensation for CO2 production, you are asked if you want to use a constant RQ value, or a previously computed channel containing VCO2 (note that VCO2 must be in units of ml/min, or you may get extremely inaccurate results), or CO2 concentration data (in % deflection from ambient CO2 concentration).  If you elect to use VCO2 or %CO2 , you will be asked to select the channel containing CO2 data.  For maximal precision, be sure to use the lag correction function to synchronize the timing of oxygen and CO2 data (especially important if there are large short-term fluctuations in gas concentrations).

Similarly, if you are calculating CO2 production, you will always need to compensate for oxygen consumption (since animals rarely produce CO2 unless they are also consuming O2) and you are asked if you want to use a constant RQ value, or oxygen concentration data (in % deflection from ambient O2 concentration), or a previously computed channel containing VO2 in units of ml/min.

In most cases, use of a constant RQ will produce quite good accuracy.

Back to top

Units:  As a last step, you need to select the units for gas exchange.  In the final window of the sequence, you are presented with 4 or 5 edit fields containing mass, B.P., Ta, etc.  Adjust these as necessary.  You can select from a variety of conversion units in the pop-up menu (shown at right with ml/min selected).

If you are calculating VO2 or VCO2 in watts or joules/time, you also need to set the correct value for heat equivalency (an example for a VO2 calculation is shown at right).  The defaults (20.1 J/ml for VO2 and 25.0 J/ml for VCO2 ) are reasonable for mixed diets.  Theoretical heat values are shown for pure carbohydrate, fat, and protein diets; see Gessaman and Nagy (1988, Physiol. Zool.  61:507-513) for a discussion of gas exchange conversion factors for metabolic calculations.

You can click buttons to select appropriate values for carbohydrate, lipid, or protein substrates.  It is possible to directly enter your own value into the edit field, however.

You can show the conversion equation by clicking the 'show equation' button:

Back to top


Conversion equations and options

Equations used for computing exchange rates are derived in part from Depocas and Hart (1957; J.  Appl.  Physiol.  10:388-392), Hill (1972, J.  Appl.  Physiol.  33:261-263) and Withers (1977, J.  Appl.  Physiol.  42:120-123); others I derived myself.  If you want to see a (relatively) simple example of deriving an equation for VO2, look at this page.

The following symbols are used: FR = flow rate, V = exchange rate for the gas in question (oxygen, CO2, or water vapor), STP = factor for converting to standard conditions of temperature and pressure, Fi = input fractional concentration, Fe = excurrent fractional concentration, RQ = respiratory quotient.

 Note: if you have already done the STP correction to the a fixed flow rate or flow rate data in a separate channel -- for example, if you measured flow with a mass flow controller with STP-corrected output, or if you used the STP converter in the OUTPUT menu and saved the results-- make sure the temperature and pressure are set to 0 °C and 760 torr, respectively.

ONE MORE CAUTION:  the algorithms used here are appropriate for most -- BUT NOT ALL -- respirometry systems.  Check to make sure your system exactly matches the conditions outlined below.

For VO2, the calculations depend on the position of the flowmeter and whether (and where) water vapor and CO2 are absorbed.  Flow rate (FR) through the metabolism chamber is best measured upstream from the chamber in most cases.  If you need to humidify the air stream, the best arrangement is to use an upstream flow meter, with the air dried before it enters the flow meter and then re-humidified between the flow meter and the chamber.  For accurate calculations it is usually necessary to remove water vapor and helpful to scrub CO2 from the air stream downstream from the animal chamber, before O2 content is measured.

 If both CO2 and water are scrubbed and the flowmeter is upstream from the chamber (Mode 1):

VO2 = STP * (FiO2 - FeO2) * FR / (1 - FeO2)


If both CO2 and water are scrubbed and the flowmeter is downstream (Mode 3; Mask mode 1 uses the same equation after correction of flowrate for incurrent water vapor content):

VO2 = STP * (FiO2 - FeO2) * FR / (1 - FiO2)


These are the simplest and most accurate ways to compute VO2

For some flow configurations you need to know RQ for accurate calculation of VO2 , or you must have a previously computed channel containing VCO2 in ml/min, or have data on %CO2 in excurrent air (expressed as the difference from incurrent CO2 concentration).  If the animal is metabolizing carbohydrate, RQ is 1.0; if it metabolizes fat the RQ is about 0.7; if it metabolizes protein the RQ is about 0.8.  Mixed diets yield intermediate RQ.  If you don't know RQ or diet, the potential error is minimized if you use the default RQ of 0.85 (see Gessaman and Nagy 1988 and especially for estimates of energy equivalence: Koteja 1996, Functional Ecology 10, 675-677).  Also, you should be aware that there are situations where the CO2/O2 ratio can exceed 1.0 (for example, if the animal is depositing fat).

CAUTION:   If you are using the 'calculate from %CO2' or 'calculate from VCO2' modes, you need to exactly synchronize the oxygen and CO2 channels in time.   This is especially important in serial configuration (first read CO2, then read O2) because the two analyzers are not 'looking' at the gas stream simultaneously (the lag correction option in the EDIT menu can usually fix this problem).   Note that if your subject's metabolism is changing rapidly, this may not be possible even if you split your sample gas stream and read O2 and CO2 in parallel, as the response times of different gas analyzers are usually unequal.   For example, the response time of a typical CO2 analyzer is seconds, while some O2 analyzers that use high-temperature zirconium cells (like those from Applied Electrochemistry) respond in milliseconds.

 In 'calculate from constant RQ' mode:

If the flowmeter is upstream and CO2 is not removed from the excurrent air stream at any point (Mode 2):

VO2 = STP * (FiO2 - FeO2) * FR / (1 - FeO2 * (1 - RQ))


If the flowmeter is downstream and CO2 is not scrubbed from the excurrent air stream prior to flow measurement but IS removed prior to oxygen measurement (Mode 4 or Mask Modes 2 and 4):

VO2 = STP * (FiO2 - FeO2) * FR / (1 - FiO2 + RQ * (FiO2 - FeO2))


If the flowmeter is downstream and CO2 is not removed from the excurrent air stream at any point (Mode 5 or Mask Modes 3 or 5):

VO2 = STP * (FiO2 - FeO2) * FR / (1 - FiO2 * (1 - RQ))

 In 'calculate from VCO2' mode:

If the flowmeter is upstream and CO2 is not removed from the excurrent air stream at any point (Mode 2):

VO2 = STP * ((FiO2 - FeO2) * FR - FeO2* VCO2) / (1 - FeO2)


If the flowmeter is downstream and CO2 is not scrubbed from the excurrent air stream prior to flow measurement but IS removed prior to oxygen measurement (Mode 4 or Mask Modes 2 and 4):

VO2 = STP * ((FiO2 - FeO2) * FR + VCO2 * (FeO2 - FiO2)) / (1 - FiO2)


If the flowmeter is downstream and CO2 is not removed from the excurrent air stream at any point (Mode 5 or Mask Modes 3 or 5):

VO2 = STP * ((FiO2 - FeO2) * FR - FiO2 * VCO2)/ (1 - FiO2)

 In 'calculate from % CO2' mode:

If the flowmeter is upstream and CO2 is not removed from the excurrent air stream at any point (Mode 2):

VO2 = STP * FR * ((FiO2 - FeO2) - FeO2* (FeCO2 - FiCO2)) / (1 - FeO2)


If the flowmeter is downstream and CO2 is not scrubbed from the excurrent air stream prior to flow measurement but IS removed prior to oxygen measurement (Mode 4 or Mask Modes 2 and 4); note that this mode is sensitive to whether CO2 is scrubbed from the incurrent air stream:

VO2 = STP * FR * ((FiO2 - FeO2) - FiO2 * (FeCO2 - FiCO2) + FeO2 * FeCO2) / (1 - FiO2)


If the flowmeter is downstream and CO2 is not removed from the excurrent air stream at any point (Mode 5 or Mask Modes 3 or 5):

VO2 = STP * FR * ((FiO2 - FeO2) + FiO2 * (FiCO2 - FeCO2))/ (1 - FiO2)

Back to top


For VCO2, you need to know RQ, or you must have data on oxygen content or consumption.  If you don't know RQ or diet, use the default RQ of 0.85, which usually produces results with an error of <5%.  Alternately, you can use a previously-computed channel containing VO2 in ml/min, or have data on %O2 in excurrent air (in difference from incurrent O2 concentration).  The equations (which assume gas is dry when FeCO2 is measured) are as follows:

CAUTION:   If you are using the 'calculate from %O2' or 'calculate from VO2' modes, you need to exactly synchronize the oxygen and CO2 channels in time.   This is especially important in serial configuration (first read CO2, then read O2) because the two analyzers are not 'looking' at the gas stream simultaneously (the lag correction option in the EDIT menu can usually fix this problem).   Note that if your subject's metabolism is changing rapidly, this may not be possible even if you split your sample gas stream and read O2 and CO2 in parallel, as the response times of different gas analyzers are usually unequal.   For example, the response time of a typical CO2 analyzer is seconds, while some O2 analyzers that use high-temperature zirconium cells (like those from Applied Electrochemistry) respond in milliseconds.

 In 'calculate from constant RQ' mode:

If the flowmeter is upstream from the chamber (Mode 1 or Mask Mode 1):

VCO2 = STP * (FeCO2 - FiCO2) * FR / (1 - FeCO2 * (1-(1/RQ)))


If the flowmeter is downstream from the chamber and CO2 is not scrubbed prior to flow measurement (Mode 2):

VCO2 = STP * (FeCO2 - FiCO2) * FR / (1 - FiCO2 * (1-(1/RQ)))


If the flowmeter is downstream and CO2 is scrubbed prior to flow measurement (Mode 3 or Mask Mode 2):

VCO2 = STP * (FeCO2 - FiCO2) * FR / (1 - FeCO2 + FiCO2/RQ)


If you are using a mask and the flowmeter is immediately downstream from the animal, and water is scrubbed prior to CO2 measurement (Mask Mode 3):

VCO2 = STP * (FeCO2 - FiCO2) * FR / (1 - FiCO2 * (1 + 1/RQ))

 In 'calculate from VO2' mode:

If the flowmeter is upstream from the chamber (Mode 1):

VCO2 = STP * (FR * (FeCO2 - FiCO2) - FeCO2* VO2) / (1 - FeCO2)


If the flowmeter is downstream from the chamber and CO2 is not scrubbed prior to flow measurement (Mode 2 or Mask Mode 1 or Mask Mode 3):

VCO2 = STP * (FR * (FeCO2 - FiCO2) - FiCO2* VO2) / (1 - FiCO2)


If the flowmeter is downstream and CO2 is scrubbed prior to flow measurement (Mode 3 or Mask Mode 2):

VCO2 = STP * (FR * (FeCO2 - FiCO2) - FeCO2* VO2) / (1 + FeCO2)

 In 'calculate from % O2' mode:

If the flowmeter is upstream from the chamber (Mode 1):

VCO2 = STP * FR * ((FeCO2 - FiCO2) - FeCO2* (FiO2 - FeO2)) / (1 - FeCO2)


If the flowmeter is downstream from the chamber and CO2 is not scrubbed prior to flow measurement (Mode 2 or Mask Mode 1 or Mask Mode 3):

VCO2 = STP * FR * ((FeCO2 - FiCO2) + FiCO2* (FiO2 - FeO2)) / (1 + FiCO2)


If the flowmeter is downstream and CO2 is scrubbed prior to flow measurement (Mode 3 or Mask Mode 2):

VCO2 = STP * FR * ((FeCO2 - FiCO2) + FiCO2* (FiO2 - FeO2)) / (1 + FeCO2 - FiCO2)

Back to top


For Evaporative Water Loss, the program assumes there is no gas volume change between the part of the system where flow rate is measured (whether upstream or downstream of the chamber) and the point where water content is measured.  The equation is:

EWL = (FeH2O - FiH2O) *STP * FR

with appropriate correction for whatever output units are desired.  This ignores any effects of VO2, VCO2, and changed water vapor content on flow rate (in combination, these exchanges usually have a very minor effect on calculated EWL).  The equation is simple, but figuring out FiH2O and FeH2O from typical humidity data is not, because of the non-linear relationship between temperature and the water content of a gas. 

Water vapor content is usually measured as either percent relative humidity or dew point temperature (in °C).  Select the appropriate units for your humidity sensor's output.  In either case, values must be converted into water vapor densities and then into fractional concentrations.  The equations for computing water vapor density are arithmetically rather nasty (and therefore the speed of conversion isn't as fast for some EWL calculations as for other kinds of gas exchange).  If you use % RH, the algorithms need to know the temperature of the humidity sensor.  You can either enter this directly or the value can be obtained from a data channel.  In the latter case, and for all analyses of dew point data, vapor density calculations must be repeated for each sample point.  That slows the rate of conversion considerably, so a progress bar is shown.

The algorithms used to compute vapor density are derived from Properties of Air, by Tracy, Welch, and Porter (1980; University of Wisconsin; you can find a pdf on the Web via Google Scholar).  In turn, these are based on the Smithsonian Meteorological Tables.   For those interested the formulae used are as follows:

• Vapor pressure (pw) at temperatures over liquid water (Smithsonian Tables, 1984, after Goff and Gratch, 1946):

Log10 pw = -7.90298 (373.16 T-1)
                    + 5.02808 Log10(373.16 / T)
                    - 1.3816 10^-7 (10^11.344 (1-T / 373.16) -1)
                    + 8.1328 10^-3 (10^-3.49149 (373.16 / T-1) -1)
                    + Log10(1013.246)
with T in °K and pw in hPa

• Vapor pressure (pi) at temperatures below 0 °C (over ice) (Smithsonian Tables, 1984):

Log10 pi =   -9.09718 (273.16/T - 1)
                    - 3.56654 Log10(273.16/ T)
                    + 0.876793 (1 - T/ 273.16)
                    + Log10(6.1071)
with T in °K and pi in hPa

The Goff-Gratch equation (for air over liquid water) covers a temperature range of -50 °C to about 100 °C, but is mostly theoretical for very low temperatures.   Accuracy is probably ±0.5% or better at temperatures between -20 and 70 °C.

• CAUTION: regardless of the accuracy of the equations, technically it is very difficult to avoid condensation or freezing of water vapor when working at subzero temperatures.

Back to top

Miscellaneous notes for gas exchange:

The 'instantaneous' correction helps compensate for the volumetric washout characteristics of respirometry systems, using the 'Z' correction approach modified for exponential washout kinetics (Bartholomew, Vleck, and Vleck 1981, Journal of Experimental Biology 90, 17-32).  It works best in systems with large flow rates relative to effective volume.  In general, it is prudent to avoid this somewhat rude manipulation of data if the system volume is large and the flow rate is low, unless the chamber is very well stirred (as in a recirculating wind tunnel or a chamber with mixing fans).  In particular, be aware of the following:

NOTE:  Movements of animals within chambers, relative to the position of input and output ports, can introduce very ugly errors in instantaneous calculations because of transient artifactual enrichments or rarifactions of local gas concentrations that have nothing to do with real metabolic rates.

Basically, the instantaneous conversion algorithm works from estimations of how rapidly a gas concentration that has been transiently deflected from a constant value will return to a state of zero deflection if no other changes occur.  Another way of putting it is as follows: if an animal instantly changes its metabolism to a new steady-state value, how long does it take for the excurrent gas concentrations to achieve their new steady state values? An instantaneous change in metabolism will be detected at the gas analyzer as a gradual approach to new steady-states, as shown schematically here:

What the 'instantaneous' conversion does is back-calculate from the measured concentration changes to approximate the real event:

For VO2 calculations the algorithm assumes incurrent oxygen concentration is 20.95% and offset so as to read zero at 20.95% oxygen.  For carbon dioxide and water vapor it assumes that incurrent concentrations are zero % (not offset).  Calculations will yield inaccurate results if these assumptions are violated.  The so-called effective volume (below) is necessary for these calculations.  It can be derived from a washout curve for the system being used.

The response correction option in the TRANSFORMATIONS routines (EDIT menu) can also adjust for mixing problems and washout characteristics.  It is based on an unmodified Z-transform, and in some cases may be easier to use than the instantaneous correction.

Back to top


  •   SENSOR HUMIDITY COMPENSATION...  Normally, the preferred procedure is to scrub water vapor from the sampled gas stream before measuring the concentrations of O2 or CO2.  However, in some respirometry systems it is necessary or desirable to analyze 'wet' gas -- i.e., gas that contains some water vapor.  For example, you might want to make uninterrupted measurements over a very long period.  This introduces a potential problem: any desiccants used to scrub water vapor may be exhausted before the measurements are completed (and you won't be able to tell which data were obtained from dry gas and which were not).  By not removing water vapor at all, you can avoid that complication.

    Note: this is related to, but not the same as, the effect of water vapor on measurements of flow rate, as discussed on this page.

    In order to accurately compute gas exchange rates from undried gas, it is necessary to remove the dilution effect of water vapor in the gas stream (which slightly reduces the concentration of O2 or CO2 in the gas analyzer).  This causes an underestimate (VCO2) or overestimate (VO2) of 'true' gas exchange rates.  The resulting error is small if the temperature and relative humidity of the measured gas are low, but increase rapidly as temperature and humidity climb (see the figure in the section on water vapor correction showing the volumetric contribution of water vapor at different temperatures and humidities).  Low barometric pressure increases the error still further.

    You use several methods for compensating for measurements on wet gas, but for all of them you must provide information of the temperature and water content of the measured gas stream.  You can use recorded channels of relative humidity or dew point, or provide a constant value.  Similarly, you can use recorded or constant values for temperature.  These are selected through a series of windows.  Note that some combinations require extensive vapor pressure calculations for each data point, which slows the computation speed; in these cases a progress bar is shown.

    Note that these routines compensate only for the passive dilution effect of water vapor.  They cannot rectify any active interference with the sensor caused by water.  In other words, if a sensor detects water vapor and 'confuses' it for the gas of interest (CO2 or O2), there is no alternative to completely drying the gas stream prior to measurement. Fortunately, most O2 and CO2 analyzers are relatively immune to this problem.

    Entering the required data involves several successive windows, the first of which is shown at right.  You need to choose between VO2 or %O2 deflection (the difference between breathed and inbreathed air) versus VCO2, %CO2, or absolute %O2 content (i.e., the 'raw' concentration of O2 prior to zeroing to baseline).   This selection is necessary because water vapor in the sensor dilutes the concentrations of other gases.   If you calculated VO2, or zeroed O2 content relative to reference, that dilution will artifactual increase the calculated VO2 (because it increases the difference between the O2 content of dry, unbreathed air and humid breathed air).   However, for VCO2, or %CO2, or unadjusted O2 content, water vapor tends to decrease the calculated VCO2 (or VO2) because it reduces concentrations.   This seems confusing, but remember than animals consume oxygen (i.e., lowering O2 content of air going to the sensor) but produce CO2 (i.e., raising CO2 content of air going to the sensor).   Finally, you can select if you want to store the result in a new channel (if the number of channels is less than the maximum channel count).

    Next, you need to specify the type of humidity measurement, which can be relative humidity, dew point temperature, vapor pressure (kiloPascals), or vapor pressure (micrograms/mL).   At this stage, indicate whether you are using vapor pressure or RH or dew point.

    Note that it's usually necessary to adjust to ambient barometric pressure and ambient temperature (even if flow rate has already been set to STP).  This is because the partial pressure of water vapor is constant at a given combination of relative humidity and temperature, regardless of the partial pressure of the rest of the gas mixture.   You also have the option of not compensating, if you wish.

    You are next asked for a channel containing humidity (or water vapor) data; alternately you can use a constant RH (the default is 20%).  This is shown below, at left.  Next you are next asked whether the humidity data are in units of relative humidity or dew point temperature or (if using vapor pressure) if the units are kiloPascals or micrograms/mL (below, at right):

      

    If the data are in units of relative humidity, you are asked to select a channel containing humidity temperature data (as shown at right), or a constant temperature for the RH sensor: 

    Once all these data are entered, the software will compensate for temperature differences between the humidity sensor and the flowmeter.

    Back to top


  •   VO2 IN WATER...   This routine will calculate oxygen depletion curves or rates of O2 consumption of aquatic organisms, based on changes O2 concentration. The latter can be expressed in a variety of units: oxygen partial pressure, i.e., pO2 (torr), percent saturation of oxygen (zero to 100%), parts per million O2 (by mass), or mg O2/Liter (note that the latter two units are identical).

    • For closed-system measurements, LabAnalyst estimates the amount of dissolved oxygen in the container at the start of measurements from container volume, temperature, solute concentration, initial oxygen saturation, and the partial pressure of O2 in the gas phase (calculated from barometric pressure and the percentage of O2 in the gas mixture, and accounting for the water vapor content of air at 100% RH). 
    • For open-system measurements, LabAnalyst estimates the amount of dissolved oxygen in the water flux through the system from temperature, solute concentration, initial oxygen saturation, and the partial pressure of O2 in the gas phase (calculated from barometric pressure and the percentage of O2 in the gas mixture, and accounting for the water vapor content of air at 100% RH).  Flow rates are obtained either from a channel or from keyboard entry (you can also specify whether deflections in oxygen content are positive or negative). 

    Most of these parameters can be edited with the EDIT FILE DATA option, including flow rate, barometric pressure, temperature, and container volume (= effective volume).  Alternately, you can enter your own O2 content data (dissolved oxygen per unit volume) and click the 'net O2' button, or you can directly enter net O2 content (total dissolved oxygen).

    If there are less than 40 channels in the file, you can store results in a new channel.

    For closed-system measurements, there are three conversion options.  For any of these, you can generate absolute units or mass-specific units (per mg, per g, or per kg).  Note that all conversions are based on the value shown in the 'net O2 content' edit field.  If you know that oxygen content differs from the computed value in this edit field, you can enter your own value prior to doing the conversion. 

    • cumulative oxygen content. This simply converts the input data (% saturation, pO2, ppm, mg/L) into the residual oxygen content of the container.  You can select several different output units, such as mL, microliters, mmoles, or micromoles.  The slope of this line over time (multiplied by -1) is the average rate of oxygen consumption (VO2). 
    • total oxygen used.  This option computes the cumulative amount of oxygen used during the measurement, again with a choice of output units.  The slope of this line over time is the average rate of oxygen consumption.  As for the previous option, several choices of units are available. 
    • oxygen consumption rate. This computes the rate of oxygen consumption (VO2; in a choice of units) by taking the point-to-point derivative of the relationship between oxygen use and time.  Note: unless the input data have been highly smoothed, results may be noisy when averaged over short time intervals.  However, long-term VO2 averages should be quite accurate, provided the O2 consumption of the organism was stable over time. 

    Here is an example showing conversion into consumption rate (ml O2 / [g .  min]):

    The Open-system (flow-through) routine will only compute oxygen consumption rate, in your choice of units.  You can use either of two algorithms, selected with buttons in the initial window:

    • In absolute value mode, input data are assumed to be absolute values, and VO2 is calculated by subtracting them from baseline O2 concentration (i.e., the amount of O2 in the incurrent water stream).  Baseline O2 concentration is computed from pressure, temperature, etc.  as described above.  For example, if baseline O2 concentration is 100% saturated, an oxygen saturation datum of 1% is assumed to mean that 99% of the oxygen in the water stream has been consumed, and VO2 is computed accordingly. 
    • In change from ambient mode, baseline O2 concentration must be offset to zero (this should be done with the BASELINE option prior to oxygen calculations).  In other words, an oxygen saturation datum of 1% is assumed to be a 1% change from baseline percent, not an absolute value of 1%. 

    Deflections in O2 concentration in the excurrent stream may be either positive or negative (if the latter, click the 'negative deflection' button in the initial window).

    Back to top


  •    O2 HEAT EQUIVALENCE...

  •    CO2 HEAT EQUIVALENCE...   These options let you set the caloric equivalent of one ml of oxygen or CO2; the defaults are 20.1 joules/ml O2 and 25.0 joules/ml CO2. You can click buttons to select appropriate values for carbohydrate, lipid, or protein substrates.  It is possible to directly enter your own value into the edit field, however.

    This window appears automatically during gas exchange calculations if you select output units of energy instead of gas volume.


  •    EFFECTIVE VOLUME...   (Active channel only) Uses a recorded gas washout curve to compute the 'effective volume' of an open-circuit respirometry system.  Effective volume is an expression of the ratio of system volume to flow rate (it corresponds only approximately to the real volume of chamber and plumbing).  This routine bases its calculations on washout rates of any gas, as long as the washout deflection is measured as a change in % concentration and the equilibrium concentration is offset to a 'baseline' of zero.  This is done either during measurement or with the baseline correction and transformations routines in LabAnalyst.

    To generate a washout curve, set up the respirometry system as it is normally used (but without an animal) at the flow rate used for measurements.  Make a recording of gas concentration, starting at equilibrium levels.  After establishing a baseline, deflect concentration by rapidly exhaling into the incurrent flow upstream from the chamber (or quickly inject a bolus of some gas with different concentration than reference gas).  Continue recording as gas concentration rapidly peaks, slowly declines, and eventually returns to equilibrium concentration.  Save the data.  In LabAnalyst, correct the baseline to zero, even for oxygen files (it will probably help to smooth the data also).  Mark a block in the washout curve as it returns to equilibrium.  Usually it is best to select from the middle of the washout curve.  Then select the effective volume option.  LabAnalyst will request the flow rate and compute effective volume.

    Back to top


    Flow rate measurement accuracy

    In open-system respirometry, the most frequent source of error is probably inaccuracies in measurement of gas flow rate.  Aside from the required correction to standard temperature and pressure (STP; 0 °C, 760 torr), there are several common problems that can compromise accuracy:

  •    An inaccurate calibration of the flow measuring instrument.  Most of the various devices used to determine gas flow rates -- turbine-based meters, rotometers ('ball in a tube' meters), mass flow meters, etc.  -- require periodic recalibration to maintain measurement accuracy.  Even if the initial factory calibration is dead on, there will be some drift over time.  Note that this is particularly true for mass flow devices (both meters and controllers).  After extensive use, the inevitable slight buildup of airborne contaminants on the sensor will cause calibration drift.  There may also be some gas leakage or pressure change in the reference portion of the sensor system.

    • Recalibration can be done in a number of ways.  You can send your units back to the factory (at considerable cost in time and expense).  Alternately, you can check your meter against another meter with a calibration you trust.  Better yet, you can calibrate against a volume meter (like a bubble meter or a good dry volume meter) that is relatively immune to calibration drift (be sure to correct for STP during the calibration process).
               One good do-it-yourself calibration method, particularly useful with dry-volume meters, is to evaporate a carefully weighed quantity of liquid nitrogen or dry ice (CO2) through the meter, and then checking the cumulative volume against the expected volume of the evaporated gas (22.4 liters per mole, at STP).   Be sure to account for effects of air pressure (if different from 760 Torr) and temperature (if different from the meter's calibration temperature).

  •   Working at unusual altitudes or pressures.  Most flow meters are designed and calibrated for a specific pressure range, usually fairly close to standard sea-level air pressure.  Usually, modest deviations from that standard have little effect on accuracy, but if you go to high altitude (several thousand meters) or work in strongly hyperbaric conditions, you should check your calibration.  Even some mass flow devices are not immune to this problem (although they are supposed to compensate for the effects of pressure changes).

    • The accuracy of devices that measure gas volume changes directly (bubble meters and dry volume meters) is generally unaffected by changes in working pressure (at least within the range of pressures tolerated by most air-breathing organisms). 

  • Use of unusual gas mixtures.  Most flow devices are calibrated for a specific gas (or gas mixture) and may require a correction factor when used with different gases.  Generally, for the kinds of devices used by people doing respiromentry, the factory calibration is for pure nitrogen or for air.  As an example of the effects of non-standard gas mixtures, consider 'helox' (21% oxygen, 79% helium).  This mix is often used by physiologists to induce cold stress without risk of freezing injury, or to alter the diffusion and viscosity properties of the breathing mixture (it also makes you talk in a squeaky voice if you breath it).  If you use helox with a mass flow device calibrated for nitrogen or air, the real flow rate for most such units is about 33% greater than the indicated flow (this is because mass flow devices measure flow from rates of heat transfer, and the heat transfer properties of helium are quite different from those of nitrogen).   However, I've seen considerably greater errors (2X - 3X) with some mass flow instruments.

    • Devices that measure gas volume changes directly (bubble meters and dry volume meters) are essentially unaffected by different gas mixtures. 

  • Other common sources of error

  •    An inaccurate calibration of the gas analyzer.  This is mainly an issue for humidity (water vapor) and CO2 analyzers; most modern oxygen analyzers rarely require span calibration -- in other words, a 1.0% change in O2 concentration is correctly read as a 1.0% change.   They DO require periodic reference readings during measurements to account for drift, however.  This means that even if you set the reference (unbreathed air) O2 concentration to read as the atmospheric value of 20.95%, subsequent reference checks will drift away from 20.95%.
            For humidity and CO2 sensors, even if the factory calibration is 'perfect', there will be some loss of span accuracy over time -- more with some instruments, less with others.  For both humidity and CO2 analyzers, the calibration procedure usually involves reading a gas with zero concentration of water vapor or CO2, adjusting the 'zero' setting on the analyzer, and then reading a gas with an independently known concentration of water or CO2 and adjusting the 'span' (or similar) setting.   Refer to the analyzer manual for details.   How often this needs to be done varies a lot between instruments but it should be a regular part of your laboratory routine, particularly with sensitive measurements.   I'm often surprised when researchers contact me about some issue or other, and the conversation eventually gets around to analyzer calibration, and it becomes clear they had no idea it was necessary.

    • With either water or CO2 analyzers, you need a non-zero 'span' (calibration) gas of known concentration of the gas of interest.  For CO2, one typically purchases a cylinder (or two) of a very expensive precision gas mixture, with CO2 concentration close to but slightly higher than what you expect to be produced during your animal measurements.  Of course, you're at the mercy of the vendor's procedures for the accuracy of the span gas concentration.   It's also possible to start with pure 100% CO2 and CAREFULLY dilute it with CO2-free air to a known concentration (but I don't know anybody who does this).

    • Calibrating humidity analyzers is often tricky.   For one thing, capacitance-based sensors (which are common) are usually inaccurate at very high humidities (say, > 90%), so it isn't appropriate to use saturated gas (i.e., 100% RH).   What's usually done is to measure water vapor (or humidity) over saturated salt solutions, which depress water vapor pressure compared to that over pure water. Different salt species give different saturation vapor pressures.   It's challenging to mix these in a way that guarantees 'true' saturation, but it can be done with reasonable accuracy.   A paper with some values for different solutes is Wexler & Hasegawa, Journal of Research of the National Bureau of Standards53: 19-26 (1954); another is Winston & Bates, Ecology 41: 232-237 (1960).   There are other relevant references if you search the literature; Properties of Air by Tracy, Welch, and Porter (University of Wisconsin and available on the Web via Google Scholar) has extensive tables for different solutes.

  •   Leaks from the animal chamber or elsewhere in the plumbing.  Depending on the construction of metabolism chambers, it may be difficult or impossible to make them completely leak-free.  Often leaks don't affect measurement accuracy, as long as the amount of air (or other gas) entering the metabolism chamber is accurately known. For positive-pressure ('push') systems with upstream flowmeters, leakage downstream is often unimportant, as long as enough uncontaminated sample gas gets to the gas analyzers.   For negative-pressure ('pull') systems, there cannot be leaks between the chamber, the flowmeter, and the device generating the suction.   In either context, you need to make sure that your measured gas (either directly from the animal chamber gas outflow, or sub-sampled from that outflow) is NOT contaminated with room air.   For example, if you sub-sample at 100 ml/min, but only 70 ml/min reaches the subsampling system, you've got problems.   As a general rule, I always check for leaks by blocking tubing exiting the chamber and observing the flowmeter -- if there are no leaks, flow will drop to zero (of course, if the inflow is at high pressure, you may pop a tubing connector or burst the chamber...).   I also often check the approximate metabolic rate against the expected metabolic rate (obtainable for most species from the metabolism calculator in the SPECIAL menu.
  • Back to top


    Other links: