Linking chlorophyll a fluorescence to photosynthesis for

Preparing to load PDF file. please wait...

0 of 0
Linking chlorophyll a fluorescence to photosynthesis for

Transcript Of Linking chlorophyll a fluorescence to photosynthesis for

Downloaded from at Viikki Science Library on May 27, 2014

Journal of Experimental Botany Advance Access published May 27, 2014
Journal of Experimental Botany doi:10.1093/jxb/eru191
Darwin Review
Linking chlorophyll a fluorescence to photosynthesis for remote sensing applications: mechanisms and challenges
Albert Porcar-Castell1,*, Esa Tyystjärvi2, Jon Atherton1, Christiaan van der Tol3, Jaume Flexas4, Erhard E. Pfündel5, Jose Moreno6, Christian Frankenberg7 and Joseph A. Berry8
1  Department of Forest Sciences, University of Helsinki, PO Box 27, 00014 Helsinki, Finland 2  Molecular Plant Biology, Department of Biochemistry, University of Turku, FI-20014 Turku, Finland 3  Faculty of ITC, University of Twente, PO Box 217, 7524 AE Enschede, The Netherlands 4  Plant Biology under Mediterranean Conditions, Universitat de les Illes Balears, Ctra. de Valldemossa Km. 7.5, 07122 Palma, Spain 5  Heinz Walz GmbH, Eichenring 6, D-91090 Effeltrich, Germany 6  Department of Earth Physics and Thermodynamics, Faculty of Physics, University of Valencia, C/ Dr. Moliner, 50, 46100 Burjassot, Valencia, Spain 7  Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA 91109, USA 8  Department of Global Ecology, Carnegie Institution of Washington, Stanford, CA 94305, USA
*  To whom correspondence should be addressed. E-mail: [email protected]
Received 14 February 2014; Revised 21 March 2014; Accepted 31 March 2014
Chlorophyll a fluorescence (ChlF) has been used for decades to study the organization, functioning, and physiology of photosynthesis at the leaf and subcellular levels. ChlF is now measurable from remote sensing platforms. This provides a new optical means to track photosynthesis and gross primary productivity of terrestrial ecosystems. Importantly, the spatiotemporal and methodological context of the new applications is dramatically different compared with most of the available ChlF literature, which raises a number of important considerations. Although we have a good mechanistic understanding of the processes that control the ChlF signal over the short term, the seasonal link between ChlF and photosynthesis remains obscure. Additionally, while the current understanding of in vivo ChlF is based on pulse amplitude-modulated (PAM) measurements, remote sensing applications are based on
Abbreviations: AG, gross photosynthetic rate (e.g. μmol CO2 fixed by the Calvin–Benson cycle m–2 s–1); AN, net photosynthetic rate (e.g. μmol CO2 exchanged at the leaf scale m–2 s–1); APAR, the flux of photosynthetically active radiation absorbed by plants (e.g. μmol photons m–2 s–1); aI, aII, relative absorption cross-section areas of PSI and PSII populations; range from zero to one; CET, cyclic electron transport; ChlF, chlorophyll a fluorescence; ETR, electron transport rate through PSII (μmol electrons m–2 s–1), equivalent to LET; F, fluorescence signal emanating from a leaf, as in Equation 11; fAPAR, fraction of incoming PAR absorbed by vegetation; range from zero to one; FII, FIIM, prevailing and maximal fluorescence signal at the level of PSII, respectively; FPSI(λem),function that accounts for the shape of the fluorescence emission spectra in PSI; FPSII(λem), function that accounts for the shape of the fluorescence emission spectra in PSII; FR, FFR, fluorescence measured in the red or far-red region, respectively; F′, FM′, prevailing and maximal fluorescence signal as measured with PAM fluorometry (relative units, e.g. sensor mV output); F0, FM, minimal and maximal fluorescence signal as measured with PAM fluorometry in the dark and after a period of dark acclimation (relative units, e.g. sensor mV output); ΦF, quantum yield of fluorescence (quanta emitted/quanta absorbed); ΦFI, ΦFII, quantum yield of fluorescence in PSI and PSII, respectively; ΦP, quantum yield of photochemistry in PSII (electrons transported/quanta absorbed); ΦPmax, maximum quantum yield of photochemistry in PSII obtained after dark acclimation or during the night; GPP, gross primary productivity (e.g. total μmol CO2 assimilated by plants m–2 s–1); kCII, rate constant of excitation energy transfer between neighbouring PSII units that denotes the degree of excitonic connectivity (s–1); kD, rate constant of basal or constitutive thermal energy dissipation (s–1); kF, rate constant of chlorophyll fluorescence emission (s–1); kISC, rate constant of intersystem crossing (s–1); kNPQ, rate constant of regulated thermal energy dissipation or NPQ (s–1); kP, overall rate constant of photochemistry (s–1); kPSII, intrinsic rate constant of photochemistry in open and fully functional reaction centres (s–1); kRCI, rate constant of thermal energy dissipation by closed PSI (s–1); kT, rate constant of energy transfer between neighbouring pigments (s–1); LET, linear electron transport; LUE, light use efficiency of photosynthesis (μmol CO2 assimilated/quanta absorbed); NPQ, non-photochemical quenching of excitation energy via regulated thermal energy dissipation; NPQ (in italics), a fluorescence parameter to quantify NPQ capacity, proportional to kNPQ/(kD+kF); PQ, photochemical quenching of excitation energy via electron transport; PQ (in italics), a fluorescence parameter to quantify PQ capacity, proportional to kP/(kD+kF); PR, photorespiration rate (μmol CO2 photorespired m–2 s–1); PRI, photochemical reflectance index; PSI, photosystem I; PSII, photosystem II; q, fraction of open and functional reaction centres; range from zero to one. qL, the photochemical quenching parameter based on a lake model assumption, an estimate of q; qP, the photochemical quenching parameter based on a puddle model assumption, an estimate of q; Rd, the rate of leaf-level mitochondrial day respiration (e.g. μmol CO2 respired m–2 s–1); SIF, solar-induced fluorescence (e.g. in W m–2 s–1 nm–1 s–1), also sun-induced fluorescence. © The Author 2014. Published by Oxford University Press on behalf of the Society for Experimental Biology. All rights reserved. For permissions, please email: [email protected]

Page 2 of 31 | Porcar-Castell et al.
the measurement of the passive solar-induced chlorophyll fluorescence (SIF), which entails important differences and new challenges that remain to be solved. In this review we introduce and revisit the physical, physiological, and methodological factors that control the leaf-level ChlF signal in the context of the new remote sensing applications. Specifically, we present the basis of photosynthetic acclimation and its optical signals, we introduce the physical and physiological basis of ChlF from the molecular to the leaf level and beyond, and we introduce and compare PAM and SIF methodology. Finally, we evaluate and identify the challenges that still remain to be answered in order to consolidate our mechanistic understanding of the remotely sensed SIF signal.
Key words:  Gross primary production, GPP, leaf level, photosystem II, photosystem I, PSII, PSI, photosynthesis dynamics, pulse amplitude modulation, PAM, PSII connectivity, remote sensing, solar-induced fluorescence, sun-induced fluorescence, SIF.

Downloaded from at Viikki Science Library on May 27, 2014

1. Introduction
Photosynthesis drives the global carbon cycle. Net photosynthesis can be quantified at the leaf level by monitoring CO2 exchange using chamber enclosure systems combined with infrared gas analysers (Long and Bernachi, 2003), and at the ecosystem level using flux towers and eddy covariance techniques (Goulden et al., 1996; Baldocchi, 2008). At the landscape and regional levels, gross photosynthetic CO2 assimilation, or gross primary productivity (GPP), is inferred using models and algorithms that integrate ground observations with remotely sensed data (e.g. Heinsch et al., 2006; Williams et al., 2009; Jung et al., 2011). Remotely sensed data have been extensively used to infer GPP based on the light use efficiency (LUE) model (Moneith, 1972; Kumar and Moneith, 1981; Zhao et al., 2011). In the LUE model, GPP is proportional to incoming photosynthetically active radiation (PAR), the fraction absorbed by vegetation (fAPAR), and the LUE at which absorbed radiation is used by photosynthesis:



Remote sensing has been traditionally used to estimate the first two terms of this equation (see reviews by Hilker et al., 2008; Malenovsky et al., 2009). For example, differences in surface reflectance between the red, blue, and near infrared part of the spectrum have been exploited to derive a wide range of vegetation indices to assess fAPAR, green biomass, chlorophyll content, or leaf area index (e.g. Rouse et al., 1974; Huete, 1988; Qi et al., 1994; Huete et al., 1997; Daughtry et al., 2000; Haboudane et al., 2002). Typically, vegetation indices show a strong seasonal correlation with GPP in many plant communities (e.g. grasslands, croplands, and deciduous forests), but the correlation breaks down in evergreen plant communities where seasonal changes in GPP are strongly modulated by LUE as well as fAPAR (Equation 1) (e.g. Sims et al., 2006; Garbulsky et al., 2008). To represent the dynamics of LUE, remote sensing data have been used to classify vegetation into plant functional types (PFTs). Subsequently, global GPP models combine spatially resolved PFT and other remote sensing products [e.g. PAR, fAPAR, temperature, and vapour pressure deficit (VPD)] with functions and parameters derived from flux tower observations to estimate LUE (Heinsch et al., 2006; Williams et al., 2009; Jung et al., 2011).

Although the LUE approach provides a theoretical basis for constructing and calibrating models, there are no corresponding benchmarks to evaluate model performance at large geographical scales, and model uncertainty remains high (Beer et al., 2010). The situation could be dramatically improved if a new source of data was available that captured the dynamic behaviour of photosynthesis at the relevant scale. Fortunately, photosynthesis generates an optical signal that, in addition to PAR and fAPAR, is also sensitive to LUE. This signal is chlorophyll a fluorescence (ChlF). ChlF are photons of red and far-red light that are emitted by chlorophyll a pigments nanoseconds after light absorption. Because photosynthesis and ChlF compete for the same excitation energy, ChlF carries information on LUE.
ChlF has been used for decades to elucidate the organization, function, and acclimation of the photosynthetic apparatus at the subcellular and leaf levels (see seminal reviews by Krause and Weis, 1991; Govindjee, 1995; Lázar, 1999; Maxwell and Johnson, 2000; Baker, 2008). Originally restricted to the laboratory, ChlF measurements made their move to the field with the development of the pulse amplitude-modulated (PAM) technique (an active technique that involves the use of a measuring light and a saturating light pulse), and the subsequent introduction of commercial PAM fluorometers (Schreiber et al., 1986; Bolhàr-Nordenkampf et al., 1989). PAM fluorometry has facilitated the study of the acclimation of photosynthesis in situ and helped clarify the link between ChlF and photosynthetic CO2 assimilation. Yet, despite the importance and value of PAM fluorometry, for practical reasons the technique has been restricted to the leaf level, and its applicability at the canopy and landscape levels remains unknown. To fill the gap, the field of ChlF has recently seen a new wave of developments that seek to measure ChlF from remote sensing platforms.
The remote sensing technique is based on the passive measurement of solar-induced chlorophyll fluorescence (SIF). The goal is to use the seasonal dynamics in the SIF signal measured from towers, aircrafts, and satellites as a proxy of photosynthesis (Grace et al., 2007; Hilker et al., 2008; Meroni et al., 2009; Rascher et al., 2009). During the last decade, SIF has been successfully measured from tower (Moya et al., 2004; Rossini et al., 2010; Guanter et al., 2013; Drolet et al., 2014), aircraft (Zarco-Tejada et al., 2009, 2012, 2013), and satellite platforms (Guanter et al., 2007; Joiner et al., 2011; Frankenberg et al., 2011; Guanter et al., 2012), the

Fluorescence and photosynthesis for remote sensing  |  Page 3 of 31

Downloaded from at Viikki Science Library on May 27, 2014

latter yielding the first global maps of terrestrial ChlF from GOSAT, SCIAMACHY, and GOME-2 (Frankenberg et al., 2011; Joiner et al., 2011, 2013). New space missions, which are in the later stages of development (e.g. NASA OCO-2, GOSAT-2, and ESA Sentinels 4–5), will provide better coverage of SIF data and open up new study possibilities, by using sensors primarily designed to monitor atmospheric chemistry. In turn, the FLEX mission (Moreno et al., 2006), in the latest stage of evaluation by the European Space Agency (ESA), has been specifically designed and optimized to map ChlF at a spatial resolution of 300 m, providing high resolution and global coverage. The amount, quality, and spatiotemporal coverage of SIF data are rapidly increasing. However, we are left with important questions. Are we ready to exploit all the information carried by the SIF signal? Can we export the knowledge obtained from short-term PAM studies to decipher the seasonal dynamics in SIF?
Remote sensing of ChlF takes place at a different spatiotemporal domain and uses a different methodology compared with the majority of ChlF studies in the literature. Amongst others, PAM fluorescence is measured over a broad spectral region whereas SIF is estimated within very narrow spectral bands. PAM fluorescence is not affected by ambient illumination whereas the SIF signal is. Most importantly of all, while we have a good mechanistic understanding of the processes that control the ChlF signal over the short term (from seconds to days), the interplay between the seasonal acclimation of photosynthesis and the ChlF signal remains unknown. Clearly, there is an urgent need to compile and re-examine the underlying theory in the context of the new applications, to identify the open questions, and to establish a roadmap that encourages the needed breakthroughs.
The goal of this review is to introduce and revisit the physical, physiological, and methodological factors that control the ChlF signal in the context of remote sensing applications, and to identify those research questions that remain open. This review is also conceived as a general introduction of ChlF for the broad community involved in the remote sensing of ChlF.
For simplicity, we focus on the leaf level because it is the smallest spatial scale at which fluorescence and photosynthetic CO2 uptake can be mechanistically linked and measured simultaneously. Up-scaling the signal from the leaf to the scales observed by airborne or spaceborne sensors falls in the domain of canopy–atmosphere radiative transfer, something equally essential to interpret SIF but outside the scope of the present review.
The review is organized in four main sections. In Section 2, we present the basis of photosynthesis and its optical signals to clarify the potential and limitations of optical data, and to introduce the multitude of processes that are embedded in Equation 1. In Section 3 we introduce the biophysical and physiological basis of ChlF from the molecular level to the leaf level and beyond, providing the theoretical and mechanistic knowledge needed to understand the spatiotemporal dynamics of the ChlF signal from the context of remote sensing. In Section 4 we introduce and compare PAM and SIF fluorometry, clarifying the main differences. Finally, in Section

5 we identify and discuss the challenges that remain to be solved, proposing further experimental work. Alternatively, the reader may wish to skip the background theory presented in Sections 2 and 3 and come back to if later on for reference.
2. The regulation of photosynthesis and its optical signals
Photosynthesis involves two main sets of reactions: the light reactions, where electromagnetic energy is absorbed by pigments and converted into chemical energy in the form of ATP and NADPH; and the carbon fixation reactions, where ATP and NADPH are used to produce sugars from atmospheric carbon dioxide. Because the light and carbon reactions exhibit different sensitivities to environmental variables such as light, temperature, or water availability, the production of ATP and NADPH by the light reactions and consumption of these metabolites by the carbon reactions do not always match (Ögren et al., 1984; Huner et al., 1996; Ensminger et al., 2006).
Energy absorbed in excess by the light reactions can damage the photosynthetic machinery (Barber and Andersson, 1992; Demmig-Adams and Adams, 2000; Tyystjärvi, 2013), for example a leaf in a sunny (high energy input) but cold (low energy consumption) environment. Accordingly, plants have evolved a number of regulatory mechanisms to adjust the energy balance between the light and carbon reactions (Walters, 2005; Demmig-Adams and Adams, 2006). The result of this continuous adjustment is that the performance of the light reactions of photosynthesis (visible to optical sensors) tends to emulate that of the carbon reactions. This establishes a link between optical data and GPP that can be implemented to the remote sensing of photosynthesis. Some processes, however, interfere with the relationship, which we discuss below (Fig. 1).
2.1 Light absorption and its regulation
Photosynthesis starts with absorption of light, mainly by chlorophyll molecules. Accordingly, an effective mechanism used by plants to regulate light absorption, or fAPAR, consists of adjusting the concentration of chlorophyll pigments in the leaf (Fig. 1). The relationship between chlorophyll content and light absorption is non-linear because the increment in light absorption per unit of chlorophyll decreases at high chlorophyll contents (Adams et al., 1990; Gitelson et al., 1998). Net changes in leaf-level chlorophyll are visible over time scales of days (García-Plazaola and Becerril, 2001; Lu et al., 2001). In addition, certain plant species use other mechanisms to modulate photosynthetic light absorption that operate at different temporal scales: leaf movements and leaf angle adjustments (Yu and Berg, 1994; Arena et al., 2008), chloroplast movements (Brugnoli and Björkman, 1992; Sarvikas et al., 2010), changes in surface reflectance mediated by salt bladders (Mooney et al., 1977; Esteban et al., 2013), changes in leaf epicuticular wax properties (Pfündel et al., 2006; Olascoaga et al., 2014), changes in leaf

Page 4 of 31 | Porcar-Castell et al.

surface structures such as pubescence (Ehleringer et al., 1976; Morales et al., 2002; Galmés et al., 2007a), and changes in the concentration of non-photosynthetic pigments such as anthocyanins (Close and Beadle, 2003; Pfündel et al., 2006; Merzlyak et al., 2008) (Fig. 1). The temporal dynamics of these processes need to be considered when interpreting ChlF data because changes in light absorption have a direct impact on ChlF intensity (see Section 3). A special case is that of non-photosynthetic pigments which do not contribute to ChlF or photosynthesis, but increase leaf absorptance (Hlavinka et al., 2013).

Fig. 1.  Photosynthetic energy partitioning at the leaf level and the light use efficiency model (GPP=PAR fAPAR LUE). Red valve symbols indicate the action of regulatory mechanisms which adjust the energy partitioning between pathways (grey arrows). Grey arrows represent the flow of energy. Optical signals available to remote sensing include properties of reflected light (shown in green) and chlorophyll a fluorescence (shown in red). LET, linear electron transport; CET, cyclic electron transport.

2.2 Linear and cyclic electron transport and energy distribution between photosystems
Photosynthetic pigments are bound by proteins to form photosynthetic antenna complexes (Liu et al., 2004) that capture light energy and transfer it to a reaction centre. A reaction centre is a special pigment–protein complex that converts excitation energy to chemical energy. The combination of reaction centre and antenna is termed a photosystem. Higher plants have two types of photosystems: photosystem I (PSI) and photosystem II (PSII), which actually operate in series in the opposite order, that is with electrons being transferred from PSII to PSI (Fig. 2).
After a photon is captured by a chlorophyll molecule in PSII, the excitation energy rapidly reaches the reaction centre chlorophyll, referred to as P680 (a pigment with absorption maximum at 680 nm). Excited P680* rapidly gives an electron to the primary electron acceptor, pheophytin, which in turn reduces the quinone A (QA) electron acceptor to yield the first stable charge-separated state P680+QA–. Subsequently, QA– passes an electron to quinone B (QB) which leaves its binding site when double reduced and protonated by stromal protons.

Downloaded from at Viikki Science Library on May 27, 2014

Fig. 2.  Schematic representation of the photosynthetic linear electron transport (LET) chain. In LET, excitation energy from absorbed photons in PSII is
used to reduce the plastoquinone pool (‘PQ’) and pump protons from the chloroplast stroma into the thylakoid lumen via the Cyt b6f complex. Energy from absorbed photons in PSII is also used to operate the oxygen-evolving complex (OEC) by which water molecules are split yielding further protons. Simultaneously, the energy from photons absorbed by PSI is used, via ferrodoxin (Fd), by ferrodoxin-NADPH reductase (FNR) to reduce NADP+ to NADPH. The oxidized PSI reaction centre P700+ is reduced back to P700 using an electron donated by plastocyanin (PC), originally from PSII. In cyclic
electron transport (CET), an electron is passed from Fd back to the PQ pool and again to PSI via PC. This results in pumping of protons to the lumen but
no NADPH synthesis. Protons accumulated in the lumen (from either LET or CET) are used by ATP synthase (ATPase) to synthesize ATP.

Fluorescence and photosynthesis for remote sensing  |  Page 5 of 31

Downloaded from at Viikki Science Library on May 27, 2014

Protonated QB reduces plastoquinone (‘PQ’, not to be confounded with photochemical quenching PQ), which is subsequently reoxidized by the cytochrome b6f complex (Cyt b6f), eventually transferring the protons to the thylakoid lumen (Fig. 2). At the donor side, P680+ is reduced by tyrosine Z (TyrZ); subsequently TyrZ+ takes up an electron from the oxygen-evolving complex (OEC) which is responsible for the splitting of the water molecule and the release of oxygen and protons (Antal et al., 2013). The resulting protons, together with those pumped to the lumen by the Cyt b6f complex, accumulate in the thylakoid lumen, generating a proton concentration gradient across the thylakoid membrane (Kramer et al., 2004a). This gradient is used by ATP synthase (ATPase) to synthesize ATP. Simultaneously, energy absorbed in PSI is captured by its reaction centre chlorophyll (P700) and used to reduce the electron acceptor ferrodoxin (Fd). Oxidized P700+ is reduced back to P700 by taking an electron from plastocyanin (PC). From ferredoxin, the electron is passed to NADP+ to produce NADPH in a reaction catalysed by ferrodoxinNADPH reductase (FNR). This series of reactions makes up the linear electron transport (LET) (for reviews, see Ort and Yocum, 1996; Antal et al., 2013).
Efficient operation of the LET implies that the populations of PSII and PSI work in series and their reaction centres transfer electrons at approximately similar rates. Although a trivial solution would be to allocate the same relative antenna crosssection area to PSII and PSI, extra flexibility is required. For example, the absorption spectra of PSII and PSI are different due to differences in pigment composition and spectral forms (see Section 3.2), with PSI absorbing light of slightly longer wavelengths (Duysens and Sweers, 1963; Boichenko, 1998; Pfündel, 2009). In particular, PSI has a higher proportion of chlorophyll a compared with PSII, with chlorophyll a/b ratios of 9 for PSI compared with 2.5 for PSII, in extracted photosystem particles (Ben-Shem et al., 2003; Nield and Barber, 2006). As a result, the probabilities of light absorption by PSII and PSI will change depending on the spectral properties of incoming light, and its temporal and spatial dynamics.
Another factor that calls for extra flexibility in energy partitioning between photosystems is the operation and dynamics of cyclic electron transport (CET) (Joliot and Joliot, 2002; Rumeau et al., 2007). CET translocates electrons around PSI and pumps protons from the chloroplast stroma to the thylakoid lumen (Fig. 2). Because the electron is recycled, CET yields only ATP but no NADPH. Functionally, CET is thought to contribute to the efficient induction of the Calvin–Benson cycle upon illumination of dark-acclimated leaves (Joliot and Joliot, 2002), the regulation of the lumen pH, and thereby modulation of non-photochemical quenching in the photosystems, NPQ (see below) (Kramer et al., 2004a), or the protection of PSI against photoinhibition (Rumeau et al., 2007; Sonoike, 2011). Indeed, CET has been found to be essential for normal carbon fixation in many C4 plants (Hatch, 1992). To summarize, changes in light quality and CET require flexible mechanisms capable of adjusting the energy partitioning between PSII and PSI (i.e. the relative absorption cross-sections of PSII and PSI, aII and aI, respectively) (Fig. 1).

At a time scale of minutes, the partitioning of energy between photosystems (aII and aI) is regulated through a process known as state transitions (Murata, 1969; Haldrup et al., 2001; Tikkanen et al., 2011). Under low light, if leaves are illuminated with light that favours PSII, part of the peripheral antenna complexes of PSII can migrate to serve PSI. This has the effect of balancing the energy input between the photosystems. State transitions are considered to be important only under low light conditions (Rintamäki et al., 1997; Haldrup et al., 2001; Tikkanen et al., 2011), and are therefore of little relevance for remote sensing.
At time scales of days, photosystem stoichiometry and aII and aI can adjust in response to more sustained changes in light intensity and quality (Anderson et al., 1988; Chow et al., 1990; Durnford and Falkowski, 1997; Pfannschmidt et al., 1999; Haldrup et al., 2001). Light intensity and quality co-vary within a plant canopy, with shaded parts of the canopy or understorey plants receiving light enriched in farred due to absorption of red light by foliage above. The result is that leaves in shaded environments tend to display higher aII:aI ratios than more exposed foliage (Anderson et al., 1988; Chow et al., 1990; Rivadossi et al., 1999; Hihara and Sonoike, 2001; Eichelmann et al., 2005; Ballotari et al., 2007). In addition to their functional and energetic role, the spatiotemporal dynamics of aII and aI have important implications for the interpretation of ChlF because they affect the shape of the fluorescence spectra and the magnitude of the PSI fluorescence contribution (Palombi et al., 2011), as well as the estimation of the LET rate by means of fluorescence (see Section 4.1).
Evidence suggests that CET and energy partitioning between photosystems might be highly dynamic in response to stress and environmental conditions (Martin et al., 1978; Ivanov et al., 2001; Eichelmann et al., 2005; Rumeau et al., 2007). However, the seasonal and spatiotemporal dynamics of these traits remain poorly understood.
2.3 Energy partitioning at the photosystem level
Understanding the processes that control the energy partitioning in PSII is crucial to linking ChlF with photosynthetic CO2 assimilation. Energy absorbed by pigments of PSII is dissipated by three main pathways: (i) it can be used by photochemistry (by LET); (ii) it can be dissipated nonradiatively as heat; or (iii) it can be re-emitted as a photon of fluorescence. A unique relationship between ChlF and photochemical efficiency cannot be established. This is because non-radiative dissipation of excitation energy is dynamic and under physiological control (see Section 3).
In low light and in the absence of stress, most of the absorbed energy is effectively used by photochemistry, and the excitation lifetime in the antenna of PSII (τPSII) is short (in the order of hundreds of picoseconds) (Dau, 1994; Lavergne and Trissl, 1995; Gilmore et al., 1995). This results in lowered fluorescence yield. This de-excitation pathway is termed photochemical quenching (PQ) (a term originally coined to denote the quenching of the fluorescence signal but herein used to address the photochemical quenching of excitation

Page 6 of 31 | Porcar-Castell et al.

Downloaded from at Viikki Science Library on May 27, 2014

energy). If light intensity increases, the carbon fixation reactions and electron transport chain gradually become light saturated, causing an increase in τPSII. This results in increased fluorescence yield. The sudden increase in fluorescence yield, observed when subjecting a dark-acclimated leaf to strong illumination, and the subsequent decrease are collectively referred as the Kautsky effect (e.g. Govindjee, 1995). These rapid fluorescence dynamics reflect the rapid reduction and re-oxidation of PSII electron acceptors and their influence on τPSII (see Section 3.2) (Brody and Rabinovitch, 1957).
Plants are incentivized to keep τPSII as low as possible, while still permitting photochemical trapping. This acts to minimize the formation of chlorophyll triplet states which might lead to production of singlet oxygen, a hazardous reactive oxygen species (Barber and Andersson, 1992). As a result, plants have evolved a number of regulatory mechanisms that are capable of dissipating the excess quanta as heat (see reviews by Müller et al., 2001; Demmig-Adams and Adams, 2006; Garcia-Plazaola et al., 2012). The operation of these mechanisms results in a decrease in the excitation lifetime in the antenna and subsequently a decrease in the ChlF yield. This type of de-excitation pathway is commonly referred to as non-photochemical quenching (NPQ), after being widely estimated using the fluorescence parameter NPQ (Bilger and Björkman, 1991) (see Section 4.1). However, it should be kept in mind that the regulated thermal dissipation of excitation energy (non-photochemical quenching of excitation energy) and the parameter NPQ (non-photochemical quenching of the fluorescence signal) are not always equivalent; thus, it becomes practical to separate NPQ and NPQ (Porcar-Castell 2011; García-Plazaola et al., 2012). For example, when state transitions reduce the PSII relative absorption cross-section (aII), the fluorescence signal is lowered (and the NPQ parameter increases) because fewer photons are absorbed by PSII (Horton and Hague, 1988), but no change takes place in τPSII. In the following, we use NPQ and NPQ accordingly.
The regulation of NPQ in PSII involves mechanisms operating at different time scales. In the short term (seconds to hours), two ΔpH-dependent mechanisms appear to regulate thermal energy dissipation in PSII (Müller et al., 2001; Demmig-Adams and Adams, 2006; Garcia-Plazaola et al., 2012). When the electron transport chain saturates, proton accumulation tends to decrease lumen pH (Fig. 2). Subsequently, the PsbS protein acts as a proton sensor activating and de-activating NPQ, while lumen pH regulates the activity of the enzyme violaxanthin de-epoxidase (VDE). The lowering of the pH triggers the de-epoxidation of violaxanthin to zeaxanthin, resulting in amplified NPQ (Demmig-Adams, 1990; Horton et al., 1996; Müller et al., 2001; Jahns and Holzwarth, 2012). This second mechanism operates at time scales of minutes. Together, the protonation of antenna proteins and de-epoxidation of xanthophyll cycle pigments have been traditionally addressed as energy-dependent quenching (qE) (Weis and Berry, 1987; Krause and Weis, 1991; Horton et al., 1996). Recently, another zeaxanthin-dependent (qZ) but ΔpH-independent form of NPQ was found in Arabidopsis and suggested also to operate at a time scale of minutes (Nilkens et al., 2010). All these mechanisms modulate NPQ

over the course of the day in response to diurnal fluctuations in light and temperature. These NPQ forms relax in the dark (e.g. overnight) and are accordingly termed flexible or reversible NPQ (Müller et al. 2001; Demmig-Adams et al., 2006; Porcar-Castell, 2011).
Over longer time scales (days to weeks) plants face more sustained changes in their environment (e.g. drought or low winter temperatures). At the seasonal scale, the light-harvesting machinery undergoes sustained changes that result in down-regulation of the photochemical quenching capacity (PQ) and up-regulation of the non-photochemical quenching capacity (NPQ) in PSII (Ottander et al., 1991, 1995; Verhoeven et al., 1996; Ensminger et al., 2004; Porcar-Castell et al., 2008a; Porcar-Castell, 2011). These adjustments are termed sustained because they do not recover overnight. The decrease in PQ is associated with the accumulation of damaged/photoinhibited PSII reaction centres. Damage and recovery of reaction centres take place simultaneously (Kok, 1956; Ohad et al., 1984; Greer et al., 1986). Thus photoinhibition becomes apparent when damage occurs faster than recovery and, since the rate constant of photoinhibition is proportional to incoming light intensity (Tyystjärvi and Aro, 1996) while the rate of recovery is temperature dependent (Greer et al., 1986), photoinhibition becomes apparent under strong light and particularly when strong light is combined with low temperatures (Strand and Lundmark, 1987; Ottander et al., 1991; Campbell and Tyystjärvi, 2012; Tyystjärvi, 2013). The term photoinhibition has also been used to denote a decrease in the maximum quantum yield of photochemistry, commonly estimated via the fluorescence parameter FV/FM (see Section 4.1). The seasonal decrease in FV/FM or ‘photoinhibition’ has been shown to be caused by reaction centre damage, the presence of sustained NPQ, or a combination of both (Porcar-Castell et al., 2008b). In this review, we use the term photoinhibition to refer to the damage of the reaction centre exclusively.
The increase in sustained NPQ has been associated with the overnight retention of a de-epoxidized xanthophyll cycle, the accumulation of the PsbS protein, the aggregation of lightharvesting complexes (LHCs) (Adams and Demmig-Adams, 1994; Ottander et al., 1995; Verhoeven et al., 1996; Ensminger et al., 2004; Zarter et al., 2006), as well as with changes in the redox properties of the electron acceptors of PSII that promote thermal dissipation in the reaction centre (Krause, 1988; Ivanov et al., 2002, 2008; Matsubara and Chow, 2004); see Verhoeven (2014) for a recent review on sustained forms of NPQ. While the correlation between NPQ (i.e. thermal energy dissipation) and the parameter NPQ (i.e. fluorescence signal quenching) is well understood over the diurnal scale, the correlation between NPQ and NPQ at the seasonal scale remains obscure (Porcar-Castell, 2011), which in turn complicates the interpretation of seasonal time series of ChlF data.
In contrast to PSII, the lifetime of excitation in PSI (τPSI) does not seem to be affected by photochemical and non-photochemical quenching processes (at least over the short term) possibly because its reaction centre is very efficient in quenching excitation energy in the oxidized state. The result is a relatively low and constant contribution of fluorescence from

Fluorescence and photosynthesis for remote sensing  |  Page 7 of 31

Downloaded from at Viikki Science Library on May 27, 2014

PSI to the total signal (Genty et al., 1990b; Pfündel, 1998; Palombi et al., 2011) that can be treated as a constant signal offset (e.g. Porcar-Castell et al., 2006). However, the seasonal dynamics of PSI ChlF remain unknown.
In addition to ChlF, the regulation in energy partitioning at PSII generates another optical signal. The operation of the xanthophyll cycle and the fast regulation of NPQ in PSII are associated with changes in reflectance at ~531 nm (Gamon et al., 1990). This feature is exploited by the photochemical reflectance index (PRI) (Gamon et al., 1992), an index that has been shown to track changes in LUE through its correlation with NPQ (Evain et al., 2004; Nichol et al., 2006). The potential of the PRI as a remote sensing proxy of LUE has been demonstrated (Drolet et al., 2005; Garbulsky et al., 2008). However, the PRI has also been shown to be very sensitive to canopy structure, gap fraction, background, viewing angle, or leaf area index (Barton and North, 2001; Sims et al., 2006; Goerner et al., 2011), complicating the association of PRI and LUE. In addition, although the short-term variation in leaf-level PRI appears indeed to be controlled by NPQ, the seasonal variation in leaf-level PRI seems to be controlled by the slow changes in pigment pools rather than NPQ (Stylinski et al., 2002; Filella et al., 2009; Porcar-Castell et al., 2012). The mechanistic link between the PRI and LUE appears to be highly dependent on scale and remains to be fully elucidated.
2.4 Alternative electron transport sinks and metabolic pathways
ChlF can be used to estimate the rate of LET through PSII (see Section 4.1). However, a number of processes need to be taken into account if we are to infer the rate of gross photosynthetic CO2 assimilation or the rates of ATP and NADPH production from fluorometric estimates of LET. Because CET produces ATP but no NADPH, fluorometrically estimated LET will decouple from ATP production in the presence of CET (Fig. 1). Additionally, alternative electron transport sinks such as chlororespiration (Nixon, 2000) and the Mehler reaction (Asada, 2000) reduce the overall quantum yield of NADPH, further decoupling LET from NADPH production.
Alternative electron sinks are generally assumed to be small relative to LET, although they play an important functional role and can become significant under certain conditions. For example, the rate of chlororespiration has been shown to increase under high light and high temperature (Diaz et al., 2007) and suggested to contribute to excess energy dissipation in the alpine plant Ranunculus glacialis (Laureau et al., 2013). The Mehler reaction, in turn, has been shown to be stimulated under some water stress conditions (Biehler and Fock, 1996; Flexas et al., 1999; Asada 2000). In addition, part of ATP and NADPH produced by the light reactions can be used by alternative metabolic pathways (e.g. nitrate and sulphate reduction in chloroplasts, or emission of plant volatile organic compounds) (Fig. 1). Krivosheeva et al. (1996) found that the electron transport rate and photosynthetic CO2 uptake were decoupled in overwintering Scots pine, suggesting an increase in an alternative electron sink or metabolic pathway. Overall, the action of alternative sinks and metabolic pathways may,

under certain conditions, affect the ability of ChlF to track the dynamics of photosynthetic CO2 assimilation.

2.5 Carboxylation, oxygenation, and day respiration

The ATP and NADPH generated by the light reactions are utilized by the Calvin–Benson cycle to synthesize sugars by assimilating CO2 (gross photosynthetic assimilation or AG) (Fig. 1). Net photosynthetic assimilation (AN) is the quantity that is measurable by gas exchange systems and relates to ‘true’ or gross photosynthesis (AG) as:

AN = AG – PR – Rd


where Rd is the rate of mitochondrial day respiration and PR is the rate of photorespiration (Ogren, 1984). In photorespiration, Rubisco catalyses the oxidation (adding O2) instead of the carboxylation (adding CO2) of ribulose bisphosphate, with the oxidized product being partially recovered via the emission of CO2. Because fluorometrically estimated LET (see Section 4.1) relates to AG rather than AN, the magnitude of photorespiration and mitochondrial respiration needs to be taken into account when comparing gas exchange and ChlF measurements.
In C4 plants (e.g. maize and sorghum), photorespiration is almost fully suppressed, and fluorometrically estimated LET correlates well with net photosynthesis (e.g. Genty et al., 1990a; Krall and Edwards, 1992). In C3 plants, which include practically all tree species and the majority of higher plants, the energy flow going to photorespiration is considerable and variable. Flexas and Medrano (2002) showed that for a pool of different species the quantum yield of photorespiration increased in response to mild water stress, from 18% to 22%, while the quantum yield of photochemistry stayed approximately constant. Thus, processes such as photorespiration may undermine the capacity of ChlF to track plant stress under certain conditions.
The temperature sensitivity of mitochondrial respiration may vary seasonally, diurnally, or within a plant, in response to variations in maintenance respiration (Atkin et al., 2005). For example, Rd has been found to increase from 15% of AN to as much as 50% in response to drought stress (Flexas et al., 2005; Galmés et al., 2007b), and a similar temporal pattern has been observed in boreal Scots pine foliage, with higher Rd values during spring recovery of photosynthesis compared with summer (Kolari et al., 2007). In summary, CET, alternative electron sinks, photorespiration, and mitochondrial respiration can all decouple optical data such as ChlF from net photosynthetic CO2 assimilation.

3. Physical and physiological controls of chlorophyll a fluorescence across space and time
In this section we describe how the intensity, spectrum, and dynamics of the ChlF signal in vivo are controlled by a number of scale-dependent physical and physiological factors (including photosynthesis).

Page 8 of 31 | Porcar-Castell et al.

Fluorescence is radiative loss of the energy of absorbed photons. Because part of the energy of the absorbed photon is lost as heat, the energy of the emitted photon is usually lower (longer wavelength) than that of the absorbed photon, a phenomenon known as Stokes shift. The quantum yield of fluorescence emission depends on both the properties of the fluorescing chromophore and its surroundings. Chlorophylls in ether are highly fluorescent, with quantum yields of 30% and 15% for chlorophyll a and chlorophyll b, respectively (Latimer et al., 1956; Barber et al., 1989). Carotenoids, in turn, yield very little fluorescence (Gillbro and Cogdell, 1989). In contrast, the quantum yield of ChlF in vivo does not exceed 10%, with typical values under steady-state illumination of 0.5–3% (Latimer et al., 1956; Brody and Rabinovitch, 1957; Krause and Weis, 1991). This drastic decrease is due to the photochemical and non-photochemical quenching of excitation energy in the photosynthetic antennae, making ChlF such a valuable tool for assessing photosynthesis.
3.1 Chlorophyll fluorescence at the molecular level
Energy levels of atomic and molecular orbitals are quantized, but molecules present wide, continuous absorption and emission (fluorescence) spectra because vibrational energies of the molecule are superimposed on each electronic energy level.

A photon can be absorbed if its energy equals the difference in the sum of electronic, vibrational, and rotational energies between an excited-state orbital and the ground-state orbital. Vibrational energy is vibration of atoms about their equilibrium positions in the molecule. Because photosynthetic pigments in vivo are tightly packed into a protein matrix, the vibration of the chemical bonds in the pigment–protein complex add additional variability to the absorption and emission spectra of pigments in vivo. This phenomenon, known as inhomogeneous broadening, explains why photosynthetic pigments display different spectral forms in vivo (Vassiliev and Bruce, 2008) but not when isolated. Rotational energy is associated with the rotation of the molecule around its axis, which is negligible in the solid state.
In Fig. 3 we portray an idealized representation of the energy levels and possible energy dissipation pathways for a chlorophyll a molecule embedded in a photosynthetic antenna. Chlorophyll a absorbs photons of blue and red light very efficiently. Absorption of a blue photon raises an electron from the ground-state orbital (S0) to an orbital of a high excited state. Part of the energy is transformed to molecular vibrations and rapidly dissipated (Jennings et al., 2003). Subsequently, because vibrational energy levels from higher excited states overlap with those of the S1 level, electrons originally in an orbital of a higher excited state can switch to

Downloaded from at Viikki Science Library on May 27, 2014

Fig. 3.  Idealized Jablonski diagram illustrating the energy partitioning of absorbed photons in a chlorophyll a molecule. Upon absorption of a photon of
blue light, an electron from the ground state is raised to a higher energy state. The energy is rapidly dissipated non-radiatively as heat mainly by internal
conversion, and the electron rapidly relaxes to the first excited state (S1). In contrast, absorption of a red photon produces the S1 state directly. It is from S1 that photosynthetic energy partitioning starts. The electron can relax to the ground state via emission of a chlorophyll fluorescence photon (with an associated rate constant kF), via non-radiative thermal dissipation, by means of either constitutive (kD) or physiologically regulated mechanisms (kNPQ), via energy transfer to another pigment or to the reaction centre chlorophyll (kT), or via intersystem crossing to form a chlorophyll triplet state (kISC). In turn, triplet states can be deactivated via phosphorescence, reaction with oxygen to form singlet oxygen, or by transfer to a carotenoid.

Fluorescence and photosynthesis for remote sensing  |  Page 9 of 31

Downloaded from at Viikki Science Library on May 27, 2014

the vibrational manifold of the S1 level, and the vibrational energy is again rapidly dissipated. The phenomenon is called internal conversion and it populates the first excited state of a chlorophyll molecule in hundreds of femtoseconds to 10 picoseconds after absorption of a photon (Dau and Sauer, 1996; Jennings et al., 2003; Clegg, 2004). A photon of red light produces the S1 state directly. It is important to note that the difference in energy between a photon of red light and a photon of shorter wavelengths (higher energy) is virtually always dissipated as heat; for this reason, it is more practical to express photosynthetic efficiencies on a quantum rather than an energy basis.
Energy partitioning takes place from S1. Excitations in the S1 state live a thousand times longer compared with those in the higher energy states, and therefore other processes compete with internal conversion for dissipation of excitation from S1 to S0 (Fig. 3). In addition to internal conversion, the excitation energy can be transferred to another pigment. Excitation energy can also be lost via emission of a fluorescence photon. Furthermore, a change in the spin state can occur and produce a chlorophyll triplet state, a process called intersystem crossing (Clegg, 2004). The triplet state is long lived and relaxes either by reacting with oxygen to produce reactive singlet oxygen (Tyystjärvi, 2004), by transfer of energy to a carotenoid or other prenyllipid, by internal conversion to the ground state, or by emission of a phosphorescence photon (Hoff, 1986). Finally, the rate constant of thermal dissipation of S1 states appears to be under physiological control at the photosystem level, at least in PSII. This regulated component of thermal dissipation corresponds to the NPQ mechanism described in Section 2.3. There is no consensus on whether NPQ is enhancement of internal conversion (e.g. by a mechanism that brings chlorophyll a molecules close to each other), enhancement of transfer of energy from chlorophyll a to zeaxanthin (followed by internal conversion producing the ground state of the carotenoid), or both (Owens et al., 1992; Holt et al., 2005; van Grondelle and Novoderezhkin, 2006). We choose to introduce NPQ already at the molecular level for consistency although NPQ would be meaningful at the level of photosystem. Macroscopically, constitutive or basal thermal dissipation (D) is defined as the minimum thermal dissipation rate obtained in the absence of stress and downregulation. Any increase in thermal dissipation on top of that is attributed to NPQ.
The quantum efficiency of a process (i) that competes for excitation energy with n other processes, and where ki is the first-order rate constant of the ith process, can be expressed as (Govindjee, 2004):

∑ Φi = ki




Accordingly, the quantum yield of fluorescence at the molecular level (Fig. 3) can be expressed as:

ΦFChla =



kD + kF + kISC + kPB + kT (+kNPQ )

where kD is assumed to be the first-order rate constant associated with the process of internal conversion, kF is the rate constant of fluorescence emission, kISC is the rate constant of intersystem crossing leading to chlorophyll triplet states, kPB the rate constant of photobleaching or destruction of the chlorophyll molecule, kT the rate constant of energy transfer to a neighbouring pigment molecule, and kNPQ the rate constant of regulated thermal energy dissipation (NPQ). All reactions are of the first order and therefore the rate constants are expressed as 1/time unit (s–1). The lifetime of excitation is the inverse of the sum of all rate constants (τ=1/Σki). The rate constant of fluorescence (kF) depends on the properties of the chlorophyll molecule and is assumed to remain constant in physiological processes (Butler and Kitajima, 1975; Clegg, 2004). The rate constant kD is invariable by definition (i.e. any variation is embedded in kNPQ). The rate constants kISC and kPB are orders of magnitude smaller than kD or kF (Clegg, 2004; Santabarbara et al., 2007) and, since they are not relevant for energy partitioning, we hereafter consider them as part of kD. Finally, the rate constant of energy transfer between pigments kT includes two main mechanisms: coherent energy transfer, and Förster or fluorescence resonance energy transfer (FRET) (Nedbal and Szöcs, 1986; Clegg, 2004; Engel et al., 2007; Novoderzhkin and van Grondelle, 2010). In general, energy transfer between closely coupled pigments within a single antenna protein can be described as movement of delocalized excitons (coherent energy transfer), whereas energy transfer between pigment–protein complexes requires consideration of localized excited states (Förster energy transfer) (Novoderzhkin and van Grondelle, 2010).
3.2 Chlorophyll fluorescence at the photosystem and thylakoid membrane level
Interpreting ChlF dynamics at the level of thylakoid membrane requires the use of a number of assumptions the validity of which depends on the temporal context of the application. The assumptions listed below are generally accepted when interpreting slow ChlF dynamics (seconds to minutes), but they are too simplified to interpret fast fluorescence kinetics successfully (picosecond to second range) (e.g. Lazár, 1999; Zhu et al., 2005; Stirbet, 2013). Similarly, some of the assumptions may be equally challenged at the seasonal scale, something that will be addressed in this section.
Assumption (A): single pool model. Photosynthetic antennae have evolved to collect photons and effectively deliver the energy to a reaction centre (Fig. 4). Efficient energy transfer from pigment to pigment and to the reaction centre is very rapid (time constants 10 fs to 10 ps) (Clegg, 2004; Engel et al., 2007; Novoderzhkin and van Grondelle, 2010) in comparison with the 200–500 ps required for stable charge separation (Roelofs et al., 1992; Vassiliev and Bruce, 2008). A large body of evidence supports the idea of rapid excitation equilibration in the antennae of PSII and PSI (Schatz et al., 1988; Croce et al., 1996; Dau and Sauer, 1996; Andrizhiyevskaya et al., 2004; Miloslavina et al., 2006). If we assume rapid excitation equilibration, the antenna, core, and reaction centre can be treated as a single pigment pool

Page 10 of 31 | Porcar-Castell et al.

Downloaded from at Viikki Science Library on May 27, 2014

Fig. 4.  Chlorophyll a fluorescence and photosynthesis from the photosystem to the canopy level. (A) At the photosystem level, excitation energy is transferred (black arrows) between pigments from the outer antenna (OA), inner antenna (IA), core, and reaction centre (RC) where it can be used to drive photochemistry. A variable part of this excitation energy is lost as heat or re-emitted as chlorophyll fluorescence (F). At the photosystem level, the quantum yield of fluorescence largely depends on the capacity of the above photochemical and non-photochemical processes, whereas the fluorescence spectral properties largely depend on chlorophyll–protein conformation and antenna structure. (B) At the chloroplast level, populations of photosystem II (PSII) and photosystem I (PSI) cooperate for light absorption and together contribute to the resulting fluorescence signal, with quantum yield and spectral properties differing between photosystems. (C) At the leaf level, cells and their chloroplasts are arranged to optimize light absorption within the leaf. This distribution results in important wavelength-dependent light gradients within the leaf and wavelength-dependent reabsorption of fluorescence. (D) At the canopy level, strong vertical gradients in light quality and intensity produce gradients in photosystem size, PSII:PSI stoichiometry, thylakoid organization, leaf morphology, and leaf pigment concentrations. The chlorophyll fluorescence signal and its relationship to photosynthesis increase in complexity with increasing scale.

(Dau, 1994; Lavergne and Trissl, 1995). This greatly simplifies the analysis of ChlF data (Fig. 4). Assumption (B): no spillover. We consider that the populations of PSII and PSI are energetically isolated and that the rate of transfer of excitations from PSII to PSI (Kitajima and Butler, 1975b; Trissl and Wilhelm, 1993; Tan et al., 1998), known as spillover, can be assumed to be insignificant at ambient temperature. Assumption (C): no thermal dissipation by closed reaction centres. We consider that thermal energy dissipation by closed reaction centres (with P680+) is only relevant when analysing ChlF data that have been excited using intense laser sources, in which the time elapsed between absorption

of two consecutive photons by PSII is shorter than the lifetime of P680+ (Shinkarev and Govindjee, 1993). Assumption (D): no quenching by oxidized plastoquinone. Excitation quenching by oxidized plastoquinone is relatively small (Vernotte et al., 1979), with a rate constant that reaches 0.15(kf+kD) when at its maximum (Zhu et al., 2005) (i.e. equivalent to NPQ=0.15). Assumption (E): perfect connectivity/lake model. The main link between ChlF dynamics and photosynthesis dynamics originates at the level of PSII, via the photochemical reaction. The photochemical reaction is defined as the stable charge separation including reduction of the QA electron acceptor and advancement of the Kok