Leaf gas exchange is central to the analysis of photosynthetic processes and the development of more productive, water efficient and stress tolerant crops. This has led to a rapid expansion in the use of commercial plant photosynthesis systems which combine infra-red gas analysis and chlorophyll fluorescence (Chl-Flr) capabilities. The present review provides an introduction to the principles, common sources of error, basic measurements and protocols when using these plant photosynthesis systems. We summarise techniques to characterise the physiology of light harvesting, photosynthetic capacity and rates of respiration in the light and dark. The underlying concepts and calculation of mesophyll conductance of CO2 from the intercellular air-space to the carboxylation site within chloroplasts using leaf gas exchange and Chl-Flr are introduced. The analysis of stomatal kinetic responses is also presented, and its significance in terms of stomatal physiological control of photosynthesis that determines plant carbon and water efficiency in response to short-term variations in environmental conditions. These techniques can be utilised in the identification of the irrigation technique most suited to a particular crop, scheduling of water application in precision irrigation, and phenotyping of crops for growth under conditions of drought, temperature extremes, elevated [CO2] or exposure to pollutants.
atmospheric [CO2] (μmol mol−1)
[CO2] within the chloroplast envelope (μmol mol−1)
light absorbed by chlorophyll and then re-emitted as chlorophyll fluorescence
[CO2] within the internal sub-stomatal air-spaces (μmol mol−1)
leaf level evapotranspiration of water from stomatal and non-stomatal sources (molH2O m−2 s−1)
maximum fluorescence under steady state conditions in the light after exposure to a saturating pulse of light
steady state fluorescence under steady state conditions in the light
the maximum quantum efficiency of PSII determined by exposing the dark-adapted leaf to a saturating pulse of light
[CO2] compensation point where PN equals Rd and RPR (μmol mol−1)
[CO2] photo-compensation point where PN equals Rd (μmol mol−1)
boundary layer conductance over a leaf (molH2O m−2 s−1)
conductance of CO2 across the mesophyll layer (μmolCO2 m−2 s−1 bar−1)
stomatal conductance to water vapour (Gs H2O – molH2O m−2 s−1) or CO2 (Gs CO2 – molCO2 m−2 s−1)
total conductance to CO2 (molCO2 m−2 s−1)
infra-red gas analysers used in the measurement of [CO2] and water vapour
electron transport rate associated with photochemical carboxylation/oxygenation of RubisCO determined from gas exchange (μmole− m−2 s−1)
photochemical electron transport rate determined from Chl-Flr parameters (μmole− m−2 s−1)
electron transport for photorespiration using gas exchange (μmole− m−2 s−1)
maximum rate of electron transport required for ribulose-1,5-bisphosphate (RuBP) regeneration (μmole− m−2 s−1)
photosynthetically active radiation within the 400–700 nm wavelength spectrum of light (μmolphoton m−2 s−1)
the light compensation point (μmolphoton m−2 s−1)
light saturation point, the intensity of PAR where PN is no longer limited by PAR (μmolphoton m−2 s−1)
net photosynthesis (μmolCO2 m−2 s−1)
the maximum rate of PN at saturating levels of light and [CO2] (the part of the PN-Ci curve where PN levels off) (μmolCO2 m−2 s−1)
amount of photosynthetically active radiation (μmolphoton m−2 s−1)
rate of CO2 efflux in the light taken to represent mitochondrial respiration (μmolCO2 m−2 s−1)
relative humidity of water vapour within the atmosphere (%)
respiration in the dark (μmolCO2 m−2 s−1)
photorespiratory evolution of CO2 in the light (μmolCO2 m−2 s−1)
maximum quantum yield of PN derived from gas exchange expressed as (molCO2 photon−1)
quantum efficiency of photosystem II determined using Chl-Flr
ratio of stomata on either surface of the leaf
triose phosphate utilisation, a measure of phosphate limitations to the regeneration of RuBP in the PN-Ci curve (μmolCO2 m−2 s−1)
rate of carboxylation (μmolCO2 m−2 s−1)
maximum carboxylation capacity of ribulose-1,5-bisphosphate carboxylase/oxygenase (RubisCO) (μmolCO2 m−2 s−1)
rate of oxygenation (μmolO2 m−2 s−1)
leaf to air vapour pressure deficit (KPa)
instantaneous water use efficiency (μmolCO2 mmolH2O−1)
Gas exchange analysis of photosynthesis in vivo has become a central tool in gauging the performance of plants. The increased availability of commercial plant photosynthesis systems has facilitated an expansion in the use of gas exchange to analyse plant responses to global environmental changes such as drought, temperature stress, pollution (e.g. increased tropospheric ozone), elevated concentrations of atmospheric carbon dioxide ([CO2]) and biotic stresses that impact upon the sustainability of agriculture and/or natural ecosystems. We intend this review to serve as an introduction to the concepts, standard measurements and common sources of error involved in gas exchange analysis of plant photosynthetic physiology. We will endeavour to clarify the terminology utilised in this field to those unfamiliar with a phraseology that is inconsistently used and frequently oversteps into off-putting jargon. As many commercial plant photosynthesis gas exchange systems are now also capable of performing simultaneous measurement of chlorophyll fluorescence (Chl-Flr), we will provide an outline of how Chl-Flr parameters may be utilised alongside gas exchange to assess the status of photosynthesis. This work will focus on the potential application of leaf gas exchange analysis in plant phenotyping and precision irrigation to improve crop water use efficiency and drought tolerance. Quantification of the carbon and water fluxes of plants are central to phenotyping of tolerance to abiotic/biotic stresses, but also to efforts to enhance the productivity and water use efficiency of crops (Marino et al. 2014). For example, leaf gas exchange has been utilised effectively in the phenotyping of rice varieties exposed to different levels of water availability provided by line (Centritto et al. 2009) and sprinkler (Lauteri et al. 2014) irrigation systems, and in the identification of genotypes with enhanced drought tolerance (Chakhchar et al. 2017; Haworth et al. 2017b; Killi et al. 2018). Moreover, a combination of gas exchange on a single leaf alongside wider-scale imaging of a plant canopy can be effective in irrigation scheduling to maximise crop water productivity particularly in protected horticulture systems.
Background and principles of gas exchange
All photosynthesis gas exchange systems work by enclosing an entire leaf, or part of a leaf, within a chamber or cuvette. Within the chamber the quantity and quality of the photosynthetic photon flux density (PPFD: the amount of photosynthetically active radiation, PAR, arriving at the leaf within the 400–700 nm wavelength spectrum of light), the wind speed or ‘flow’ of air, the relative humidity (RH), temperature (leaf temperature may only be controlled to a degree under variable field conditions where the ambient conditions may diverge considerably from internal conditions within the leaf cuvette), and concentration of atmospheric gases can be controlled to determine the photosynthetic response of the area of leaf contained within the chamber. The difference in the concentration of [CO2] and water vapour between the ‘reference’ air flow that enters the cuvette and the ‘sample’ air flow that exits the cuvette (commonly referred to as ΔCO2 and ΔH2O) can be used to infer rates of photosynthesis, respiration and stomatal conductance using calculations based on the work of Farquhar et al. (1980) and Von Caemmerer & Farquhar (1981) that are automatically computed by commercial gas exchange systems. Values of ΔCO2 and ΔH2O are determined using infra-red gas analysers (IRGA). Carbon dioxide and water vapour absorb different wavelengths of infra-red radiation. The most recent ‘open system’ commercial photosynthesis gas exchange systems possess two pairs of IRGAs to concurrently measure the amount of CO2 and H2O in the reference and sample air streams. If the concentration of [CO2] and [H2O] within the air stream increases, a greater proportion of infra-red radiation is absorbed and less is received by the sensor of the IRGA. To control [CO2] within the reference air-stream, CO2 is removed from the air entering the plant photosynthesis system by soda lime, pure CO2 is then mixed with the reference air to a set concentration. Water vapour in the reference air stream can be regulated by being passed through desiccants (such as silica gel or gypsum based desiccants) or humidifiers (such as porous ceramic substrates that hold water).
The processes described in Equations (1)–(5) form the theoretical basis of the method of photosynthetic gas analysis and estimation of physiological parameters outlined later in this review. Photosynthetic electron transport can be determined using gas exchange and Chl-Flr; moreover, the simultaneous use of these two methods can provide valuable insights into the uptake of CO2 and the physiology of light capture and usage (e.g. Genty et al. 1989; Harley et al. 1992; Loreto et al. 1994).
Instantaneous point measurements of leaf gas exchange and chlorophyll fluorescence
Point measurements of leaf gas exchange can be performed under ambient or set conditions of [CO2], light intensity and temperature within the leaf cuvette. The use of ambient cuvette conditions (a transparent window is attached to the leaf cuvette instead of an LED light unit – this precludes the measurement of Chl-Flr alongside leaf gas exchange) in the field may be beneficial in terms of increased speed of measurement (as leaves do not need to adapt to vastly different conditions within the leaf cuvette) and extended battery life. However, variation in environmental conditions (e.g. changing cloud cover, temperature) may affect consistency between measurements; it is therefore preferable to conduct such measurements when possible at the same time of day on clear sunny days to minimise fluctuations in light intensity and temperature. In some cases, the impact of an experimental treatment will be evident under controlled cuvette settings (where all leaves are measured under identical conditions of [CO2], temperature, RH and PAR), but not when ambient measurements are performed because of heterogeneity in conditions in the field (e.g. Haworth et al. 2016). Instantaneous point measurements under set controlled cuvette conditions generally take longer, as the leaf requires time to adjust to the new conditions within the leaf cuvette. The maintenance of temperature, [CO2], RH and PPFD within the leaf cuvette when performing point measurements under set controlled cuvette conditions also places additional demands upon battery life if the plant photosynthesis system is not running from mains power.
Photosynthetic light response curves
The basis of photosynthesis involves the capture of PAR to provide the chemical energy to produce sugars (Evans et al. 1993). Photosynthetic light response curves (commonly termed PN – PAR curves) characterise the relationship between PN and increasing levels of PPFD. As PPFD is increased from zero, PN rises rapidly (Kok 1948); the PAR level where PN equals zero is termed the light compensation point (PARcomp). The relationship between PN and PPFD at low light intensity around PARcomp can be used to determine respiration and is outlined in more detail in the section ‘Respiration in the light and in the dark’, below. Leaves that are efficient in converting low levels of PPFD such as shade leaves generally have lower PARcomp than leaves adapted to intercept higher intensities of PPFD such as sun leaves (e.g. Demmig-Adams et al. 1989; Thornley 2002). At PPFD from 0 μmol m−2 s−1 to PARcomp, and often above PARcomp, PN rises linearly with increasing PPFD (usually this linear relationship holds to a PAR level of 200 to 400 μmol m−2 s−1); this is known as the maximum quantum yield as the availability of light limits the carboxylation of RuBP for PN (ΦCO2max, expressed as molCO2 photon−1). Beyond this point the relationship between PN and PAR is no longer linear; the curve becomes convex and begins to plateau where PN becomes limited by RubisCO (Ögren & Evans 1993); the stage at which the curve flattens represents the point where the availability of PAR no longer limits PN and is termed ‘light saturation point’ (PARsat) (Figure 1) (Ögren 1993).
When performing a light response curve, we recommend starting at PARsat and declining to lower intensities of PPFD. In species such as grasses that possess highly functional stomata that close rapidly in darkness (e.g. Haworth et al. 2018b), it is necessary to start at full light intensity and then decrease PAR rather than placing the leaf into darkness and then increasing PPFD (this approach will take comparatively longer as it is necessary to wait for the stomata to open fully at each PPFD stage). The order of the PPFD stages will not affect the shape of the curve or the calculation of parameters from the PN – PAR response curve. Cuvette temperature should be maintained at the growth temperature of the plant (normally a value in the range of 20 to 30°C is used). As stomata open or close with the change of PAR intensity, the leaf to air vapour pressure deficit (VPD) alters; this effect should be minimised by adjusting the humidity of the input air stream, which can be controlled automatically (within certain ranges) or manually. Commercial gas exchange systems control PAR using LEDs usually up to an intensity of 2,000–2,400 μmol m−2 s−1, approximately equivalent to light intensity on a clear sunny day. Red and blue light can influence stomatal opening, so it is necessary to use both when performing a PN/PAR curve (Mansfield & Meidner 1966; Zeiger & Hepler 1977; Sharkey & Raschke 1981; Loreto et al. 2009); although many plants also adsorb light in the yellow and green part of the spectrum (Terashima et al. 2009; Sakowska et al. 2018), so a more balanced quality of light is preferable. Standard steps for conducting a PN/PAR curve are: 2,000, 1,600, 1,200, 1,000, 800, 600, 400, 300, 200, 150, 100, 75, 50, 0 μmol m−2 s−1. In species adapted to low PAR, such as many ferns, PN will decline at the higher light intensities; this represents light inhibition and PAR levels above the PARsat point should not be used in photosynthetic response curves to [CO2]. Moreover, in certain species such as some grasses with rapid growth rates that are adapted to high light environments PARsat occurs above 2,000 μmol m−2 s−1. In these cases, a modelled value is assumed, but should be treated with a degree of caution. Lobo et al. (2013) provide an in-depth review of PN/PAR models and Microsoft Excel solvers for these models that permit rapid and easy determination of physiological parameters from a PN/PAR dataset.
Respiration in the light and in the dark
Respiration provides the energetic requirements for processes such as growth, repair, protective physiology and reproduction. Respiration involves the release of CO2 and can therefore be measured using gas exchange techniques. However, the respiratory CO2 release needs to be separated from the other processes involving CO2 outlined in Equation (3). Carbon dioxide produced by mitochondrial respiration and RPR is released into the cytosol as bicarbonate (HCO3−). As the concentration of [CO2] within the cytosol is lower than that within the sub-stomatal internal leaf air-space or the external atmosphere and CO2 released by photo- and mitochondrial respiration may be reabsorbed by PN, it is challenging to accurately measure respiration. Gas exchange analysis of respiration is complex, and the techniques involved are each based on their own assumptions and subject to limitations. Moreover, the fluxes of CO2 involved in gas exchange analysis of respiration are smaller than those involved in PN; therefore, any sources of error associated with leakage, temperature control or calibration of the gas exchange system have a much greater proportional effect when calculating respiration. The simplest way to distinguish respiration is to switch off the light in the cuvette (and the surrounding leaf/plant) and thus shut off PN and RPR. When the efflux of CO2 has stabilised (this will be expressed as a negative PN value in the gas exchange system output) it can be recorded and considered to represent respiration in the dark (Rn: the ‘n’ in the acronym refers to night-time respiration) (e.g. Peisker & Apel 2001; Tjoelker et al. 2001). However, this may not reflect the true respiration in the dark (Atkin et al. 1998; Pinelli & Loreto 2003), but is widely accepted as the quickest and most straightforward method of gauging respiration in plants using commercial gas exchange systems, particularly when working in the field.
Respiration also takes place in the light alongside PN and RPR (Equation (4)). This is termed respiration in the light (Rd: the ‘d’ in the acronym refers to day-time respiration) and is more difficult to measure than Rn because of the recycling of respiratory CO2 emission by PN (Pinelli & Loreto 2003). In this context, the Kok (1948) and Laisk (1977) methods that utilise gas exchange to estimate Rd can be considered to be estimates of Rd that effectively analyse CO2 evolution in the light (i.e. not ‘true’ Rd, as CO2 may be re-adsorbed by PN, as described above). By convention, respiration is sometimes expressed as a negative value, as in gas exchange systems it represents a loss of CO2 from the leaf (in comparison to PN that is positive and indicates an uptake or gain of CO2). However, even though respiration is recorded as a negative value by gas exchange systems, it should be considered and reported as positive, particularly when used in physiological calculations such as those used to estimate mesophyll conductance of CO2 (e.g. Equation (11)) (e.g. Harley et al. 1992).
Respiration in the light may be determined indirectly through gas exchange using the Kok (1948) and Laisk (1977) methods. The Kok (1948) method utilises the relationship between PN and PAR at low light intensities close to PARcomp. At PAR levels 0 to PARcomp, PN increases rapidly with PAR; however, above PARcomp the relationship between PN and PAR remains linear but is less steep. This disjunction in the PN/PAR relationship around PARcomp is known as the Kok effect and can be used to estimate Rd, as outlined in Figure 2(a), by extrapolation of the PN/PAR relationship above the ‘Kok effect’ to the point where it intersects with the y-axis (dashed line in Figure 2(a)) (Wang et al. 2001). The disjunction in the linear response between PN and PAR associated with the ‘Kok effect’ is caused by the elimination of light-induced suppression of respiration (Sharp et al. 1984). It is worth noting that in some cases the Kok effect does not occur and the relationship between PN and PAR is the same either side of PARcomp; in this scenario the points above PARcomp should still be used to extrapolate Rd, which should equal Rn. The Kok method can be used to determine Rd in both C3 and C4 plants (Bjorkman & Demmig 1987). The efficacy of the Kok method may be constrained by stomatal closure as PAR approaches zero, resulting in increased sub-stomatal [CO2] (Ci) (Crous et al. 2012). The effect of fluctuations in Ci can be compensated using the iterative approach of Kirschbaum & Farquhar (1987), which normally results in a slight increase in the level of Rd. However, this approach assumes that mesophyll conductance (Gm) is infinite, whereas experimental evidence suggests that this is not the case (e.g. Flexas et al. 2007b; Tholen et al. 2012). We would therefore advise against using such a correction, unless variations in Ci are comparatively large. The Kok method can be applied to plants grown at any [CO2] level (e.g. Haworth et al. 2016) or temperature (e.g. Haworth et al. 2018a). The intensity of PAR is decreased from a comparatively low level, with an increasing number of PAR steps as PN approaches the PARcomp point; normally PAR steps of 400, 300, 200, 150, 100, 75, 50, 30 and 0 μmol m−2 s−1 can be used but these should be manually adjusted in respect to PN when around PARcomp. Moreover, the latest generation of commercial exchange systems offer greater control of PAR intensity and quality (i.e. regulation of PPFD to ∼5 μmol m−2 s−1), permitting the construction of Kok PN/PAR response curves at a higher resolution that affords increased accuracy in the estimation of Rd.
The Laisk (1977) method generally results in lower estimates of Rd than those produced by the Kok method (Villar et al. 1994). This approach uses the response of PN to Ci taken at a number (usually 3–4) of relatively low levels of PAR (i.e. ≤400 μmol m−2 s−1). Where the PN-Ci curves intersect represents the point where PN is equal to RPR. The level of PN represented by this point on the y-axis therefore indicates Rd, as all the CO2 used by PN is released by RPR (see Equation (4)) (Figure 2(b)). As RPR is extremely sensitive, the PN – Ci curves undertaken as part of the Laisk (1977) protocol must be performed at the same temperature; this may preclude use of the method in the field where ambient temperatures may vary. The level of Ci at which this intersection occurs is known as the photo-compensation point (Γ*: pronounced gamma star – this value is different from the CO2 compensation point, Γ, as it does not include the effect of RPR). The photo-compensation point, Γ*, is an important parameter used in the modelling of PN and the estimation of parameters such as mesophyll conductance (Gm). When performing the measurement of Rd and Γ* using the Laisk (1977) method, it is necessary to use a minimum of three PAR levels and to preferably use four. The first PN – Ci curve should be performed at a PPFD of 300 and/or 400 μmol m−2 s−1, most likely the slope of curves at these two PPFD intensities will be identical. The CO2 concentration (Ca) used in the curve should be at ambient and sub-ambient level around the CO2 compensation point (Γ), usually 400, 300, 200, 150, 100, 75, 50 and 30 μmol mol−1. After each alteration in Ca, the levels of PN and Gs should be allowed to stabilise (normally 2–5 minutes per step) before gas exchange parameters are recorded. When each sequence of Ca steps is complete, the level of light in the cuvette should be decreased and the PN-Ci response curve repeated. Owing to the CO2 concentrating mechanism in the bundle sheath of C4 plants, it is not possible to use the Laisk (1977) method to estimate Rd in C4 species as the point where the lines of the PN-Ci curves intersect often occurs above the x-axis, where PN equals zero (i.e. the CO2 compensation point, Γ); this would result in a negative Rd value (the expression of Rd as a positive or a negative is discussed earlier in this section). Likewise, it is not possible to utilise this approach for C3–C4 (also known as C2) species with photosynthetic physiologies intermediate between C3 and C4 where CO2 is recaptured as part of a photorespiratory by-pass.
Gas exchange measurement of Rd and Γ* using the Laisk method is fairly complex and very sensitive to errors such as variations in temperature or leaks from the gaskets that seal the leaves within a cuvette (Flexas et al. 2007a; Rodeghiero et al. 2007). Many commercial gas exchange systems are unable to maintain a constant cuvette temperature under field conditions; it is therefore extremely difficult to successfully perform measurement of Rd using the Laisk method in the field (an air-conditioned laboratory is preferable). However, analysis of Rd and Rn in plants using CO2 composed of different stable isotopes of carbon (12C and 13C) suggests that in plants grown under optimal conditions levels of Rd and Rn are largely similar and that suppression of respiration in the light may be an artefact of the indirect method used to estimate Rd (Loreto et al. 2001). Under field conditions, where estimation of Rd may be impossible because of variations in ambient temperature, or impractical owing to time constraints, measurements of Rn may serve as an effective substitute. In fact, Rn measured under field conditions has been successfully used in place of Rd in the calculation of parameters such as Gm (e.g. Centritto et al. 2009; Lauteri et al. 2014) and other studies have used 0.5 Rn as a proxy for Rd (e.g. Niinemets et al. 2005).
The Kok (1948) and Laisk (1977) methods outlined above rely solely on gas exchange to extrapolate Rd. The majority of commercial gas exchange systems now possess the capacity to perform simultaneous measurement of Chl-Flr parameters alongside gas exchange permitting two estimates of the efficiency of CO2 uptake via photochemistry. The method of Yin et al. (2009, 2011) exploits gas exchange and chlorophyll fluorescence measurements of electron transport for photosynthesis at low irradiances. This method follows the protocol outlined earlier for the Kok (1948) method, but incorporates measurement of a Chl-Flr parameter after a saturating pulse of light. The Kok (1948) method assumes that the transport efficiency of electrons required PN under light conditions is constant, the Yin et al. (2009, 2011) method suggests that this is not the case and that electron transport efficiency declines as PAR increases (Figure 1(b) and 1(c)). Incorporation of the quantum efficiency of photosystem II (this is outlined in more detail in the section ‘Instantaneous point measurements of leaf gas exchange and chlorophyll fluorescence’, above) with gas exchange parameters produces a Rd value that is generally lower than that found using gas exchange alone in the Kok (1948) method. The Excel spreadsheet of Bellasio et al. (2016) can be used to calculate Rd based on the Yin method following the input of PN, ΦPSII and PAR parameters. One disadvantage of the Yin method is that typical commercial plant photosynthesis systems require the use of a smaller cuvette size when performing simultaneous measurements of gas exchange and Chl-Flr, whereas the Kok (1948) and Laisk (1977) approaches require only gas exchange parameters, and therefore can use larger cuvettes that incorporate a larger area of leaf, giving a more representative analysis of gas exchange and minimising errors associated with the analysis of small areas of leaf.
CO2 response curves
The response of PN to increasing [CO2] provides valuable insights into the biochemistry of photosynthesis (and photorespiration when combined with simultaneous measurement of Chl-Flr) and the diffusive limitations imposed on CO2-uptake (mainly at the stomata and mesophyll). The response of PN to [CO2] is normally expressed in terms of the relationship of PN to the sub-stomatal concentration of CO2 (Ci). This measurement is referred to as a PN-Ci curve (or more commonly as an A/Ci curve, as A is also widely used as an acronym for photosynthetic assimilation of CO2) and is one of the most common gas exchange analyses undertaken to determine the response of photosynthetic biochemistry to factors such as drought, [CO2], light intensity, temperature, nutrient status or exposure to pollution. The photosynthesis model of Farquhar et al. (1980) is used to calculate the parameters describing the biochemistry of CO2 uptake. Figure 3(a) shows a typical PN-Ci curve of a C3 species (Olea europaea).
The Ci level where PN is zero (i.e. PN = Rd + RPR, see Equation (3)) is known as the CO2 compensation point (Γ); this differs from the CO2-photocompensation point (Γ*) where PN equals RPR. At ‘low’ sub-ambient [CO2], the rate of PN is limited by the availability of CO2; in effect, there is insufficient CO2 to be used by RubisCO and any rise in Ci leads to increased availability of CO2 for carboxylation. This part of the curve is sometimes referred to as being CO2 (Wullschleger 1993) or RubisCO (e.g. Long & Bernacchi 2003) limited. As this part of the curve is limited by the availability of substrate, the relationship between PN and Ci is at its steepest – this slope can be used to calculate the maximum carboxylation rate of RuBP (Vcmax). At higher levels of Ci, where the curve begins to level off, the availability of CO2 as a substrate no longer limits PN; instead the rate of PN is limited by the electron transport required for the regeneration of RuBP (Jmax). The inflection point in the PN-Ci curve represents the transition from Vcmax to Jmax limited. Many studies regard the value of PN recorded in this stage of the PN-Ci curve at saturating levels of light and [CO2] to be the maximum rate of PN (PNmax) (e.g. Heath et al. 2005), whereas some studies define PNmax as the PN at PARsat under ambient [CO2] (e.g. Marshall & Biscoe 1980), highlighting the frequent lack of consistent use of terminology between studies. In the latter part of the PN-Ci curve it may be possible to observe evidence of phosphate limitations to the regeneration of RuBP in a reduction in PN (PN is limited by the availability of phosphate due to high concentrations of triose phosphate sugars). This part of the curve is characterised by the parameter triose phosphate utilisation (TPU) (Sharkey & Vanderveer 1989; Ellsworth et al. 2015).
The PN-Ci response curve is produced by altering the level of atmospheric [CO2] (Ca) within the leaf cuvette; as long as stomata are sufficiently open, this results in Ci following Ca (a Ci to Ca ratio of 0.7 is normally observed but values within the range of 0.5 to 0.8 can be found in plants grown under favourable conditions). The parameters from a PN-Ci response curve are reported at a standard temperature of 25 °C; however, it is not necessary to perform the PN-Ci curves at 25 °C as the effect of leaf temperature on PN is normally compensated in the Excel spreadsheets and R-statistics templates used to determine Vcmax, Jmax and TPU. Temperature (normally between 20 and 30 °C), VPD and a saturating PAR intensity (derived from the PN/PAR response of the plant used) should be maintained constant throughout the PN-Ci curve. To ensure that the plant is fully acclimatised to the conditions within the leaf cuvette, it is standard to initially expose the leaf to ambient [CO2] for 20–30 minutes before declining in stages to 50 ppm. In some plants with less robust photoprotective mechanisms, it may not be advisable to leave [CO2] below 50 ppm for a sustained period of time as the cuvette is a high energy environment at PARsat and at a standard temperature of 20 to 30 °C, and without the substrate for PN the leaf may incur oxidative damage as photochemistry is reduced (e.g. Daniel 1997; Durchan et al. 2001). At 50 ppm [CO2] the stomata should open fully and maximum rates of Gs be observed (e.g. Centritto et al. 2003; Haworth et al. 2013).
After full stomatal opening, levels of [CO2] should then be increased in steps with close attention to declines in Gs that may limit PN at higher Ca levels. Gas exchange parameters should be recorded when PN is relatively stable (±5%). Each step should take 2 to 5 minutes, and the stages at Ca levels above ambient should be fairly rapid. Standard Ca levels used in a PN-Ci response curve are 400, 350, 250, 150, 50, 100, 200, 300, 400, 600, 800, 1,000, 1,400, 1,800 and 2,000; however, these can be altered to suit the type of plant and/or experimental treatment under investigation. During drought stress, plants close their stomata to reduce transpirative water loss (e.g. Flexas et al. 2002; Lauteri et al. 2014). If Gs is too low, it is not possible to produce a meaningful PN-Ci curve as insufficient CO2 enters the sub-stomatal cavity (Lovelli & Perniola 2014). However, by extending the duration of the exposure to low Ca of 50 ppm for approximately 1 hour to fully open stomata before rapidly increasing Ca (and using fewer above-ambient Ca levels to minimise the impact of stomatal closure) it is often possible to remove the effect of stomatal diffusive limitations on PN, and therefore assess the true impact of drought stress on the physiological status of photosynthetic CO2 assimilation (Centritto et al. 2003). This approach appears to be effective in instances where Vcmax is unaffected by drought stress but Jmax may be reduced (Aganchich et al. 2009; Killi & Haworth 2017). However, in some cases the speed of stomatal closure may be too rapid to permit completion of a full PN-Ci response curve (Haworth et al. 2018b). Moreover, to accurately assess Gm using the curve fitting method, stomata must be open to minimise stomatal limitations to CO2-uptake. In well-watered plants, with stomata that do not respond to above-ambient increases in [CO2] (e.g. Haworth et al. 2015), it may be possible to take more time between Ca steps and/or add more Ca levels to enhance the resolution of the PN-Ci curve. Moreover, the number of Ca steps may depend upon the specific factors under investigation; for instance, if the area of study is the efficiency of carboxylation then an increased number of data points within the Ca range of 50 to 300 ppm [CO2] would be desirable. However, if the aim of the study is to assess possible phosphate nutrient limitations on photosynthetic physiology (e.g. Yang et al. 2016), then more data points in the higher Ca range such as 1,600 to 2,000 would be beneficial. The latest generation of plant photosynthesis systems can perform a rapid PN-Ci response curve by tracking the dynamic response of PN to [CO2] (Stinziano et al. 2017; Bunce 2018). In theory, this approach could significantly reduce the length of time required to characterise the photosynthetic physiology of plants. However, the effectiveness of this approach in removing diffusive limitations to PN in plants with low Gs or plants experiencing stress such as drought where diffusive limitations to CO2-uptake are significant is not yet apparent.
One of the most common sources of error associated with gas exchange measurements of PN is caused by leaks and/or diffusion of gases through the gaskets. The gaskets are usually made from neoprene rubber material, and are used to seal the upper and lower surfaces of the leaf within the cuvette. If the gaskets are not changed regularly, or the leaf cuvette is stored in the closed position, the gasket material may become compressed and no longer seal the leaf effectively. During measurements that take a long time, the gaskets may retain an impression of the leaf within their surface (sometimes referred to as a ‘memory’); this may lead to leaks during successive measurements as the gaskets will not seal subsequent leaves to the same degree. Leaks and diffusion of gases through the gaskets are a particular issue for the PN-Ci response curve as the concentration gradient of gases within and outside the cuvette are more pronounced. One method to counter this is to complete the PN-Ci response curve and then quantify the leaks by killing the leaf (usually by heating it for a short period within a microwave) and then returning the leaf to exactly the same position within the cuvette and repeating the Ca steps used in the PN-Ci curve (Flexas et al. 2007a). Correcting for the effect of leaks should increase PN at sub-ambient Ca and reduce PN at super-ambient Ca. We recommend an approach whereby the diffusion gradient between the air within and outside the cuvette is minimised by feeding the exhaust air (that should more closely match the air within the leaf cuvette than the ambient atmosphere) from the plant photosynthesis system into a supplementary gasket surrounding the primary gasket or a bag placed over the cuvette head (although depending upon the time to replenish the air within the bag, this may increase the duration of each Ca step in the PN-Ci curve) (Rodeghiero et al. 2007).
Determination of mesophyll conductance to CO2
curve fitting of the PN-Ci response curve (Ethier & Livingston 2004);
analysis of leaf gas exchange and discrimination between stable 12C and 13C isotopes (Evans et al. 1986; Lloyd et al. 1992) in the gas stream passing over a leaf (Vrábl et al. 2009; Douthe et al. 2012), recently synthesised sugars (Lauteri et al. 1997, 2014; Centritto et al. 2009) or leaf structural tissue (Lauteri et al. 1997).
All of these methodologies involve specific weaknesses and/or assumptions. It is therefore preferable to utilise two methodologies when determining Gm under experimental conditions (Pons et al. 2009). Nevertheless, it is necessary to note that all of the current methods employed to quantify Gm require leaf gas exchange parameters, and so are not truly independent of one another (e.g. Lauteri et al. 2014). These ambiguities associated with the determination of Gm may have contributed to observations that Gm alters rapidly to conditions such as [CO2] (Flexas et al. 2007b) or light (Douthe et al. 2012), or being reported as a fixed constant (Ethier & Livingston 2004). However, Gm can be conceptualised as a ‘flux-weighted quantity’ (Tholen et al. 2012), determined by the interaction of CO2 assimilation by PN, availability of CO2 within the internal sub-stomatal air-space (Sorrentino et al. 2016) and the biochemistry (Hanba et al. 2004; Flexas et al. 2006) and/or physical structure (Adachi et al. 2013) of the mesophyll layer.
In this introduction to gas exchange, we shall restrict our discussion to the techniques utilised for the measurement of Gm that do not require additional equipment (e.g. online isotope ratio mass spectrometers) and can be performed with a stand-alone plant photosynthesis system. More detailed review of techniques for the measurement of Gm can be found in Flexas et al. (2008) and Pons et al. (2009). The curve fitting approach of Ethier & Livingston (2004) allows the calculation of Gm from the PN-Ci response curve by assuming that Gm is not infinite (see Equation (9)). Mesophyll conductance can be calculated from the PN-Ci response curve using the Excel templates and R-scripts listed in the section ‘CO2 response curves’, above.
Stomatal conductance to CO2 (Gs CO2) is a standard data output of most commercial plant photosynthesis systems. Nevertheless, Gs CO2 can be calculated from the stomatal conductance to water vapour (Gs H2O) outlined in Equation (2). This approach is particularly important in the characterisation of environmental stresses such as drought where diffusive limitations to CO2-uptake are a major component in the plant response (an example is provided in Figure 4).
Stomata are the interface between the external atmosphere and the internal leaf environment. By regulating the uptake of CO2 for photosynthesis against the loss of water vapour as transpiration, stomata play a key role in maintaining leaf homeostasis (e.g. Haworth et al. 2010, 2011; Centritto et al. 2011a, 2011b). Changes in the turgor of the two guard cells determine the aperture of the stomatal pore. The importance of stomata in the adaptation of plants to their environment is well established (e.g. Heath 1950; Mansfield & Majernik 1970; Schulze et al. 1975; Woodward 1987). The examination of stomatal kinetics has become increasingly important in understanding plant responses to environmental change (Kardiman & Ræbild 2018; Gerardin et al. 2018; Haworth et al. 2018b), plant evolutionary history (McAdam & Brodribb 2012; Haworth et al. 2013; Doi et al. 2015) and the interaction of physiological control of guard cell turgor and stomatal morphology in regulating leaf gas exchange (Haworth et al. 2015, 2018c). Stomatal kinetics involves analysis of the dynamic response of Gs to a change in environmental conditions. In effect, the alteration of Gs serves as a proxy for the opening/closing of the stomatal pore. When measuring stomatal kinetics, it is important to maintain constant conditions within the leaf cuvette with the exception of the specific factor under consideration.
To assess stomatal opening it is standard to keep the plant in the dark for at least 12 hours prior to measurement, and perform the stomatal kinetic response at a time when the illumination of the plant normally begins (i.e. dawn for plants grown outdoors or coinciding with the time at which lights are switched on for plants grown under artificial illumination). The wavelength of light used in a stomatal opening kinetic response is critical, as blue initiates the pumping of ions across the guard cell membrane, while PN in the mesophyll associated with red light plays a role in maintaining stomatal opening (Mansfield & Meidner 1966; Zeiger & Hepler 1977; Sharkey & Raschke 1981; Doi et al. 2015). The latest plant photosynthesis systems contain green and white LEDs in addition to the standard red-blue. Green light has been shown to play an important role in driving efficient PN (Terashima et al. 2009) and is strongly absorbed in many green plants (Sakowska et al. 2018). A more balanced spectrum of light may strongly influence Gs during stomatal opening (Mansfield & Meidner 1966; Doi et al. 2006, 2015) (a similar effect may be observed in mesophyll conductance: Loreto et al. 2009; Pallozzi et al. 2013), and should be utilised where possible.
The reduction in Gs associated with stomatal closure following a complete cessation of illumination (i.e. a light to dark transition) or reduction in the intensity of PAR has been shown to be more effective than the stomatal opening kinetic in differentiating evolutionary plant groups (McAdam & Brodribb 2012; Elliott-Kingston et al. 2016; Xiong et al. 2018) or plants grown under different water availability (Haworth et al. 2018b). This stomatal closure kinetic approach also has the benefit of not being so closely constrained to a particular time of day to be most effective (e.g. Haworth et al. 2015) as found when characterising stomatal opening. The response of stomata to fluctuations in the light environment and the optimisation of PN to rapidly changing conditions is a major factor in improving the carbon and water efficiencies of crop plants. Such stomatal kinetic measurements to varying light quantities/qualities are becoming increasingly important to the development of more productive crop varieties (Lawson & Blatt 2014).
The kinetics of the Gs response to [CO2] (e.g. Heath & Meidner 1957; Haworth et al. 2013), separation of the leaf petiole from the supporting stem tissue as a proxy for desiccation (e.g. Brodribb & McAdam 2011) or leaf to air VPD (Brodribb & McAdam 2011; Haworth et al. 2018b) can also be measured. The response of stomata to a change in leaf to air VPD and increase in [CO2] above ambient levels requires the presence of light (Shimazaki et al. 2007). However, the response of stomata to a decrease in [CO2] below ambient levels occurs in both darkness and light (Mansfield et al. 1981). When measuring the kinetics of Gs to a change in conditions the ‘timed-recording’ or ‘auto-log’ function of the gas exchange system should be used. Stomatal conductance should be stable for at least 20–30 minutes prior to any change in conditions within the cuvette. When altering light intensity and/or [CO2] within the cuvette it is necessary to pay close attention to the effect of any change in Gs on leaf to air VPD. Any variation in leaf to air VPD should be minimised by adjusting the RH of the reference air-stream entering the leaf cuvette. These Gs kinetic measurements are often lengthy, and therefore consideration should be given to the performance of the two pairs of IRGAs within the plant photosynthesis system. If the time interval between measurements is sufficient, it may be possible to utilise each plant photosynthesis system's protocols to minimise drift between and/or zero the two IRGAs. However, if the frequency of measurements precludes this, it is necessary to match and/zero the IRGAs at the start and end of the stomatal kinetic response and then correct for the average drift between the reference and sample IRGAs over the duration of the measurement (e.g. Haworth et al. 2018b). Example Gs kinetics for [CO2], light intensity and leaf to air VPD are shown in Figure 5.
Application to plant phenotyping and precision irrigation
Plant phenotyping, or phenomics, is the study of morphological, biochemical and physiological characteristics of a plant under specific environmental growth conditions. A plant's phenotype is determined by the interaction of its genotype with the environment. The converging pressures of climate change, loss of productive land and population growth necessitate greater production of food and biofuel crops using fewer resources through precision agriculture techniques. However, the identification of more productive and stress resistant phenotypes has been constrained by the so-called ‘phenotyping bottleneck’ (Furbank & Tester 2011). Leaf gas exchange can provide direct measurement of photosynthetic CO2-uptake and water-loss in addition to detailed characterisation of the underlying photosynthetic physiology and stomatal behaviour. This information is highly valuable for the identification of crop varieties with desirable characteristics such as tolerance to drought (e.g. Centritto et al. 2009; Haworth et al. 2017a), and also in the identification of climate-resilient plants with high agro-ecological potential to make improved use of areas affected by erratic rainfall, drought and other associated environmental stresses (Chakhchar et al. 2017; Zegada-Lizarazu et al. 2018).
Despite the detailed information that leaf gas exchange and Chl-Flr analysis can provide regarding photosynthetic (e.g. Loreto & Centritto 2008; Centritto et al. 2009; Lauteri et al. 2014) and stomatal physiology (e.g. Killi et al. 2016), the techniques outlined above are time-consuming and constitute a significant component of the phenotyping bottleneck. Nevertheless, given the potential application of gas exchange to understanding the CO2 and water fluxes of plants, efforts to improve the capacity for gas exchange analysis in high throughput phenotyping are being undertaken (Bellasio et al. 2014). The use of leaf gas exchange analysis of a sub-set of plants alongside high throughput screening techniques such as spectroradiometry of a wider range of crop varieties may be effective in supplying both detailed physiological information and rapid high frequency analysis of plant status (e.g. Munns et al. 2010). Spectroradiometry indices such as the photochemical reflectance index (PRI) which measures the epoxidation state of xanthophylls (Gamon et al. 1992) have been shown to track photosynthetic parameters (e.g. Marino et al. 2014; Sun et al. 2014). The correlation of leaf gas exchange parameters such as PN or Gs with remote sensing may be effective in alleviating the bottleneck while adding value and validation to high throughput imaging techniques (Figure 6).
Gas exchange is one of the most effective methods to detect the early impact of water stress on plants; for example, a reduction in Gs may occur prior to any change in leaf water status (Zhang & Davies 1990; Tardieu & Simonneau 1998). However, due to the complexities associated with gas exchange analysis it is not a viable methodology to be utilised in long-term monitoring of plant water status to schedule irrigation. Nonetheless, in conjunction with wide-scale imaging, particularly canopy-scale reflectance and/or infra-red thermography, leaf gas exchange can be useful in developing calibration datasets (see Figure 6(c)) to enable effective irrigation scheduling to maximise the water productivity of a given crop (Jones 2004; Cifre et al. 2005; Patakas et al. 2005; Baker et al. 2007; Marino et al. 2014; Gago et al. 2015). Moreover, the detailed physiological information available from leaf gas exchange analysis can be vital in identifying the appropriate irrigation technique for a given crop. For example, partial root-zone drying is a common approach based on laboratory split-root studies where half of the root-zone receives full irrigation while the remainder is allowed to dry, thus generating a root-to-shoot signal from the dry half of the root-zone that promotes stomatal closure and enhances WUE (Davies et al. 2002; Romero et al. 2010). However, gas exchange analysis of a Tunisian olive variety indicated that root-zone drying was not suitable to its underlying stomatal physiology and reduced the water productivity of that specific variety (Dbara et al. 2016). Gas exchange analysis has also shown that improved PN through irrigation is only beneficial to crop yield during the specific period of fruit development. Supplemental irrigation during episodes of vegetative growth or ‘off-years’ is unlikely to enhance water productivity (Sun et al. 2014; Dbara et al. 2016). Exploitation of gas exchange analysis of photosynthetic and stomatal physiology can therefore enable the identification of the most efficient precision agriculture approaches to be utilised in the context of a given crop or the prevailing environmental conditions.
Leaf gas exchange and Chl-Flr have dramatically expanded our understanding of plant photosynthetic physiology and stomatal regulation. This is critical to the assessment of the carbon and water balance of plants and their environmental adaptation. However, these approaches are sensitive to errors and require expertise and care when conducting these measurements and interpreting the resulting data. We intend the above review to act as an introductory guide to these techniques and not to serve as an exhaustive assessment. We hope that this guide can offer a foundational starting point for researchers interested in simultaneous leaf gas exchange and Chl-Flr analysis towards plant phenotyping and optimisation of plant water use.
We are grateful for funding from the Agreement of Scientific Cooperation 2017–2019 between the National Research Council of Italy and Chinese Academy of Sciences, RGV-FAO Programma triennale 2017–2019 – Ministero delle Politiche Agricole e Forestali, and Regione Toscana PIF ‘Ottimizzazione ed Innovazione della Filiera Olivicola ed Accrescimento della Qualità dell'Olio Toscano’ Misura 16.2 Progetto ‘SEMIA’.