This is the accepted version of the article: Zhang, Chao; Atherton, Jon; Peñuelas, Josep; [et al.]. «Do all chlorophyll fluorescence emission wavelengths capture the spring recovery of photosynthesis in boreal evergreen foliage?». Plant, cell & the environment, Vol. 42, issue 12 (Dec. 2019), p. 3264-3279. DOI 10.1111/pce.13620 This version is avaible at https://ddd.uab.cat/record/218229 under the terms of the license 1 Do all chlorophyll fluorescence emission wavelengths capture the spring 2 recovery of photosynthesis in boreal evergreen foliage? 3 4 Running title: Spectral fluorescence in evergreens 5 6 Chao Zhang1, 2, 3, *, Jon Atherton1, Josep Peñuelas2, 3, Iolanda Filella2, 3, Pasi 7 Kolari4, Juho Aalto4, 5, Hanna Ruhanen6, Jaana Bäck7, Albert Porcar-Castell1,* 8 9 1Optics of Photosynthesis Laboratory, Institute for Atmospheric and Earth System 10 Research (INAR)/Forest Sciences/Viikki Plant Science Centre, Faculty of 11 Agriculture and Forestry, University of Helsinki, PO Box 27, 00014 Helsinki, 12 Finland. 13 2CREAF, Center for Ecological Research and Forestry Applications, Bellaterra 14 08193, Catalonia, Spain. 15 3CSIC, Global Ecology Unit CREAF-CSIC-UAB, Cerdanyola del Vallès 08193, 16 Catalonia, Spain 17 4Department of Physics, University of Helsinki, PO Box 64, 00014 Helsinki, 18 Finland. 19 5SMEAR II Station, University of Helsinki, Hyytiäläntie 124, Korkeakoski 35500, 20 Finland 21 6Natural Resources Institute Finland (Luke), Natural Resources and 22 Bioproduction, Suonenjoki Unit, Juntintie 154, 77600 Suonenjoki, Finland. 23 7Department of Forest Sciences, University of Helsinki, PO Box 27, 00014 24 Helsinki, Finland. 1 * 25 Corresponding authors: Chao Zhang, chao.x.zhang@helsinki.fi; Albert Porcar- 26 Castell, joan.porcar@helsinki.fi 27 Abstract 28 Chlorophyll a fluorescence (ChlF) is closely related to photosynthesis and can be 29 measured remotely using multiple spectral features as solar-induced 30 fluorescence (SIF). In boreal regions, SIF shows particular promise as an 31 indicator of photosynthesis; in part because of the limited variation of seasonal 32 light absorption in these ecosystems. Seasonal spectral changes in ChlF could 33 yield new information on processes such as sustained non-photochemical 34 quenching (NPQS), but also disrupt the relationship between SIF and 35 photosynthesis. We followed ChlF, functional and biochemical properties of 36 Pinus sylvestris needles during the photosynthetic spring recovery period to 37 answer; (1) how ChlF spectra change over seasonal timescales? (2) How 38 pigments, NPQS and total PAR absorption drive changes of ChlF spectra? (3) Do 39 all ChlF wavelengths track photosynthetic seasonality? We found seasonal ChlF 40 variation in the red and far-red wavelengths, which was strongly correlated with 41 NPQS, carotenoid content and photosynthesis (enhanced in the red), but not with 42 PAR absorption. Furthermore, a rapid decrease in red/far-red ChlF ratio occurred 43 in response to a cold spell, potentially relating to the structural reorganization of 44 the photosystems. We conclude that all current SIF retrieval features can track 45 seasonal photosynthetic dynamics in boreal evergreens, but the full SIF spectra 46 provides additional insight. 47 Key words: chlorophyll a fluorescence spectra, evergreen vegetation, F690, F740, 48 leaf PAR absorption, Pinus sylvestris, PSI fluorescence, fluorescence ratio, 49 sustained non-photochemical quenching (NPQS) 2 50 Introduction 51 Chlorophyll a fluorescence (ChlF) of leaves, thylakoids and photosystems has 52 been extensively used to study the organization, functioning, and acclimation of 53 the photosynthetic light harvesting apparatus (Govindjee 1995; Baker 2008; 54 Murchie & Lawson 2013; Porcar-Castell et al. 2014). ChlF can now be measured 55 within discrete wavelengths from plant canopies, forest stands, and whole 56 ecosystems with instrumentation on towers, drones, aircraft and satellites (Zarco- 57 Tejada, Catalina, González & Martín 2013; Porcar-Castell et al. 2015; Rascher 58 et al. 2015; Joiner, Yoshida, Guanter & Middleton 2016; Sun et al. 2017; Parazoo 59 et al. 2018). Widely addressed as solar-induced fluorescence (SIF) by the remote 60 sensing community, SIF opens up the study of photosynthesis at unprecedented 61 scales (Frankenberg et al. 2011, 2014; Parazoo et al. 2014; Thum et al. 2017; 62 Sun et al. 2018; Zhang, Guanter, Joiner, Song & Guan 2018; Zuromski et al. 63 2018). This capacity could not only serve to improve the current accuracy of 64 global carbon budgets (Damm et al. 2015; Quéré et al. 2018; Smith et al. 2018) 65 but yield new understanding on the responses and feedbacks between terrestrial 66 ecosystems and the environment, both critical milestones for the implementation 67 of climate change mitigation and adaptation strategies (Smith et al. 2014; IPCC- 68 SR15 2018). 69 Despite the mounting evidence of the strong link between SIF and gross primary 70 productivity (GPP) accumulating from ground, airborne and satellite platforms 71 (Guanter et al. 2014; Zarco-Tejada, González-Dugo & Fereres 2016; Migliavacca 72 et al. 2017; Sun et al. 2017; Parazoo et al. 2018; Zuromski et al. 2018; Magney 73 et al. 2019a), the physical and biological mechanisms that underpin the 74 relationship and the potential wavelength-dependent information content in the 3 75 SIF signal remain unclear (Porcar-Castell et al. 2014; Verrelst et al. 2016; 76 Wieneke et al. 2018; Yang et al. 2018). This lack of understanding, is particularly 77 acute for evergreen foliage, where the seasonal link between SIF and GPP 78 involves processes other than changes in absorbed photosynthetically active 79 radiation(APAR) which could decouple and add wavelength dependencies to the 80 relationship between SIF and GPP. 81 In contrast to pulse-amplitude-modulation (PAM) ChlF measurements which 82 integrate ChlF over a broad range of wavelengths, SIF is retrieved within narrow 83 and discrete spectral bands around the red and far-red ChlF emission peaks 84 (atmospheric oxygen absorption bands or solar Fraunhofer lines) (Alonso et al. 85 2007; Meroni et al. 2009). Accordingly, because the shape of the leaf-level ChlF 86 spectra is driven by a combination of physical, physiological and biochemical 87 factors (Magney et al. 2019b), it can be expected that the seasonal relationship 88 between SIF and GPP would depend on retrieval wavelength, especially in boreal 89 evergreen vegetation undergoing remarkable adjustments during the season. 90 A light use efficiency type model (Monteith 1972) becomes a convenient 91 theoretical framework to introduce the physical and biological factors that couple 92 (or decouple) SIF and GPP, and connect the leaf-level phenomenology to the 93 spatial scale of remote sensing. The intensity of the SIF signal emitted from a leaf 94 or a plant canopy at a given wavelength (λ) can be expressed as a function of 95 four factors: 1) the incoming PAR, 2) the fraction of that PAR absorbed by the 96 leaf or canopy (A), 3) the quantum yield of fluorescence with its associated 97 emission wavelength (ФF(λ)), and 4) a wavelength dependent escape probability 98 (fesc(λ)) which accounts for the reabsorption of predominantly red ChlF photons 99 by chlorophyll (Chl) molecules within the antenna, thylakoid, chloroplast, leaf or 4 100 plant canopy (Buschmann 2007; Porcar-Castell et al. 2014; Romero, Cordon & 101 Lagorio 2018; Yang & van der Tol 2018), as: 102 𝑆𝐼𝐹(𝜆) = 𝑃𝐴𝑅 ∙ 𝐴 ∙ Ф𝐹 (𝜆) ∙ 𝑓𝑒𝑠𝑐 (𝜆) (Eqn. 1) 103 In similar terms, leaf or canopy level GPP can be expressed as: 104 𝐺𝑃𝑃 = 𝑃𝐴𝑅 ∙ 𝐴 ∙ 𝐿𝑈𝐸 (Eqn. 2) 105 where LUE corresponds to the photosynthetic light use efficiency in moles of CO2 106 assimilated moles per mole of absorbed PAR photons. By combining equations 107 (note that A and PAR cancel out), a simple theoretical relationship that links GPP 108 to SIF can be obtained as: 1 𝐿𝑈𝐸 109 𝐺𝑃𝑃 = ∙ ∙ 𝑆𝐼𝐹(𝜆) (Eqn. 3) 𝑓𝑒𝑠𝑐 (𝜆) Ф𝐹 (𝜆) 110 In this equation, widely applied to the interpretation of remotely sensed SIF data 111 (Guanter et al. 2014; van der Tol, Berry, Campbell & Rascher 2014; Damm et al. 112 2015; Lee et al. 2015; Frankenberg & Berry 2018), physical factors (e.g. 113 structure-dependent ChlF reabsorption) are embedded in the 1/fesc(λ) term, and 114 biological factors (e.g. architecture and physiological state of the light reactions, 115 alternative energy sinks, cyclic electron transport, photorespiration) are 116 embedded in the LUE/ФF(λ) term. Clearly, any seasonal changes in the strength 117 of the factors above can potentially couple and decouple SIF and GPP (Porcar- 118 Castell et al. 2014). In addition, any seasonal changes in the leaf-level 119 wavelength properties of fesc(λ) and ФF(λ) can add wavelength dependencies to 120 the link between SIF(λ) and GPP, something that remains to be characterized. 121 Seasonal changes in leaf Chl content, although usually modest in boreal 122 evergreens (Öquist & Huner 2003; Ensminger et al. 2004; Porcar-Castell et al. 5 123 2008a), affect light absorption, and also fesc(λ) via ChlF reabsorption (Buschmann 124 2007). Similarly, structural changes at the level of chloroplast and thylakoid 125 membrane, e.g. thylakoid grana unstacking and aggregation of Chl binding light 126 harvesting complexes in overwintering evergreens (Öquist, Chow & Anderson 127 1992; Ruban, Johnson & Duffy 2012; Verhoeven 2014; Demmig-Adams, Muller, 128 Stewart, Cohu & Adams 2015; Ruban 2016) could also contribute to the 129 modulation of fesc(λ) potentially affecting the spectral properties of SIF and its 130 seasonal correlation to GPP. 131 Ф𝐹 (𝜆) is composed of at least two components: a highly dynamic component 132 which responds to photochemical (PQ) and non-photochemical (NPQ) quenching 133 of excitation energy in photosystem II (PSII) and that fluoresces in the red and 134 far-red regions (Franck, Juneau & Popovic 2002; Palombi et al. 2011), and a 135 component from photosystem I (PSI) assumed to remain stationary over the short 136 term (Genty, Wonders & Baker 1990; Palombi et al. 2011; Pfündel, Klughammer, 137 Meister & Cerovic 2013) that fluoresces predominantly in the near-infrared. 138 Accordingly, if seasonal dynamics in sustained NPQ (NPQS) (Ottander, Campbell 139 & Öquist 1995; Öquist & Huner 2003; Verhoeven 2014) quenched only the ChlF 140 components from PSII, one would expect NPQs to affect the spectral properties 141 of Ф𝐹 (𝜆). Similarly, differences in the patterns of seasonal photoinhibition of PSII 142 (Ensminger et al. 2004; Murata, Takahashi, Nishiyama & Allakhverdiev 2007) and 143 PSI reaction centres (Sonoike 2011; Huang, Yang, Hu & Zhang 2016) could also 144 affect the ChlF spectra and its spectral dependency with photosynthesis. 145 In summary, although a strong seasonal coupling between PAM ChlF and 146 photosynthesis has been widely reported for evergreen foliage (e.g. Ottander & 147 Öquist 1991; Ensminger et al. 2004; Zarter, Demmig-Adams, Ebbert, Adamska 6 148 & Adams 2006b; Soukupová et al. 2008; Kolari et al. 2014; Springer, Wang & 149 Gamon 2017), the spectral dependency of the relationship remains unresolved. 150 The objective of the study was to characterize the seasonal variation of ChlF 151 spectra for boreal evergreen Scots pine needles during the spring recovery of 152 photosynthesis. Hence, we investigated if red and far-red wavelengths of ChlF 153 were equally positioned to capture the spring recovery of photosynthesis. To do 154 so we combined long-term and continuous in situ measurements of gas exchange, 155 and PAM ChlF with repeated measurements of foliar pigment content, leaf total 156 PAR absorption and spectral ChlF, spanning the full dynamic range of variation 157 in photosynthetic capacity of Scots pine needles: from deeply downregulated 158 foliage during winter, to fully functional foliage during peak growing season in 159 summer. 160 161 Materials and methods 162 Study site and field sampling protocol 163 Measurements were conducted at Hyytiälä/SMEAR-II (Station for Measuring 164 Forest Ecosystem-Atmosphere Relations) in Southern Finland (61°51´N, 165 24°17´E, 181m a. s. l.) (Hari & Kulmala 2005), in a 52-year old Scots pine (Pinus 166 sylvestris L.) stand with a dominant height of c. 18m. The study period (24th 167 February - 20th July 2015) encompassed three biological seasons: end of winter 168 (full dormant state), spring, and first half of summer (peak growing season). 169 The study combined continuous in situ measurements of micrometeorological 170 variables, CO2 exchange and PAM ChlF across four different trees (N= 4), with 171 repeated point measurements of steady-state spectral ChlF at room temperature 7 172 and pigment analysis across five trees (previous four plus one, N=5). Spectral 173 ChlF was purposefully measured under standardized PAR so that it could serve 174 as a proxy of ФF(λ) in Eqns 1 and 3, (i.e. APAR assumed constant). To avoid 175 changing the needle cohort in the middle of the experiment, all measurements 176 were conducted in the newest cohort of needles available at the start of the study 177 period (developed in summer 2014). To avoid defoliation effects, several 178 branches from the four topmost whorls were selected per tree and used for the 179 repeated sampling of needles. Needles were sampled always before 10AM. 180 Overall, the study included a total of 19 measuring points in addition to the 181 continuous measurements (Fig. 1). 182 183 Continuous measurements of micrometeorological data and shoot CO2 exchange 184 Air temperature (°C) (Pt100 sensor) and PAR (μmol m-2 s-1) (Li-190SZ, Li-Cor 185 Inc., Lincoln, NE, USA) were measured right above the forest at 1 min intervals 186 from sensors placed in a tall mast. The CO2 exchange of pine shoots was 187 measured in top canopy shoots 30-60 times per day using a system of automated 188 dynamic chambers coupled to an infrared gas analyzer (Li-840, Li-Cor Inc., 189 Lincoln, NE, USA) (see e.g. Aalto et al. 2014 for further details). Chambers 190 remained open most of the time exposing the studied shoots to ambient 191 conditions and closed only during measurement. Fluxes were estimated from the 192 change in CO2 concentration during the first 40 s upon chamber closure (Kolari 193 et al. 2012). Shoots with fully developed needles from 2014 were debudded prior 194 to chamber installation to prevent new growth during the study period and for 195 practical reasons. Total needle area inside the chamber was measured at the 196 end of the study and used to calculate fluxes. A total of two to four chambers 8 197 (N=2-4) were used during the study period from February 24th to July 20th in 198 2015 (Fig. S1a). 199 Direct comparison of noon LUE from chamber data (moles CO2 assimilated 200 /moles PAR absorbed) can be problematic in boreal conditions due to low PAR 201 levels during the winter months. Accordingly, we decided to use here a simple 202 model of photosynthesis (the optimal stomatal control model) (Hari, Mäkelä, 203 Korpilahti & Holmberg 1986; Kolari, Lappalainen, Hänninen & Hari 2007) that 204 allows robust estimation of photosynthetic parameters under ambient CO 2 205 concentrations, low light and low temperatures (Kolari et al. 2014). Two key 206 photosynthetic parameters were estimated in a 3-day time window to capture the 207 seasonal development in LUE; (1) α (µmol CO2 / µmol PAR), the slope of a linear 208 function fitted to the photosynthetic light response with low incident PAR (<300 209 µmol m-2 s-1) before noon (Fig. S1b; Kolari et al. 2014). Under constant leaf PAR 210 absorption (Fig. S2), this parameter can be considered a measure of maximum 211 photosynthetic LUE under low light. In addition, and since our spectral ChlF 212 measurements were not conducted under low light, we also estimated (2) β (m s- 213 1), the maximum rate of light-saturated photosynthesis per unit inter-cellular CO2 214 concentration (Fig. S1c). This parameter can be considered a measure of light- 215 saturated photosynthesis at optimal temperature, low vapour pressure deficit and 216 ambient CO2 concentration. 217 We normalized α and β parameters to their mean values between May 3rd and 218 23rd (see Fig. S1d,e). This is a common procedure in time-series analysis and 219 serves to minimize the impact of systematic variability and emphasize the 220 seasonal patterns in the data, which was our goal. The normalization period was 221 selected in May as that was the time when all four chambers were simultaneously 9 222 recording and the foliage was already photosynthetically active. Normalized unit- 223 less values were then converted back to the original units by multiplying the 224 normalized values by the respective means across all four chambers. The 225 parameters α and β are used hereinafter as a measure of the seasonal spring 226 recovery of photosynthesis. 227 228 Continuous measurements of PAM ChlF in situ 229 A Monitoring PAM system (MONI-PAM, Walz GmbH, and Germany) (Porcar- 230 Castell, Pfündel, Korhonen & Juurola 2008b; Porcar-Castell 2011), equipped with 231 4 independent PAM fluorometers, was used to record the instantaneous 232 fluorescence yield (F’), the maximal fluorescence yield (F’M), incoming PAR, and 233 temperature every 30 minutes. Night F’ and F’M were assumed to correspond to 234 minimal (Fo) and maximal (FM) fluorescence and used to derive daily maximum 235 quantum yield of PSII, FV/FM after Kitajima & Butler (1975), and to calculate 236 quenching parameters NPQS and PQS after Porcar-castell (2011) as: NPQS= 237 FMR/FM -1 and PQS = FMR/Fo - FMR/FM, where FMR is the summer night reference 238 obtained for those particular needles in the absence of NPQS. A decrease in PQS 239 relative to summer levels was here interpreted in terms of photoinhibition of 240 reaction centres (Porcar-Castell 2011). Finally, seasonal changes in PAM based 241 fluorescence yield (ФF) were estimated as ФF= 0.1 Fo/FMR (Porcar-castell 2011, 242 note the unfortunate typo in Eqn. 24 therein where F’M should be FMR), which is 243 based on the assumption of a maximum fluorescence yield of 10% for PSII 244 particles at the FM state (Barber, Malkin & Telfer 1989). The four fluorometers 245 were installed in top canopy branches pointing south. 10 246 Validation of ChlF measurements 247 For practical reasons, this study compared ChlF spectra of cut needles measured 248 under standard conditions and at room temperature, with photosynthetic 249 parameters (α, β, FV/FM, NPQS and PQS,) obtained from the field instrumentation. 250 Accordingly, we wanted to test that our measurements at room temperature 251 correctly represented the same physiological state as measurements carried out 252 in the field. In addition, we wanted to ensure that our MONI-PAM system (which 253 consistently supplied 48 light pulses per day to the same needles throughout the 254 study period) was not introducing any long-term artefact. We used a Hansatech 255 fluorometer (FMS-2, Hansatech Instruments Ltd., Kings Lynn, Norfolk, UK) to 256 measure FV/FM both in the field and under standardized conditions indoors. 257 During each sampling point, we placed a total of 25 dark-acclimation clips, 258 distributed across the five study trees, and dark acclimated for at least 1 hour 259 prior to measuring FV/FM. The same measurements were subsequently repeated 260 at room temperature using collected needles instead. As expected, FV/FM 261 obtained from the FMS-2 in the field was highly correlated with estimates from 262 the MONI-PAM system (r=0.97, P<0.0001) (Fig. S3), denoting that the MONI- 263 PAM system did not introduce any bias into the observed seasonal patterns. The 264 small discrepancy between slopes is very likely due to differences in the colour 265 of the measuring light as well as the ChlF detection range between instruments. 266 Similarly, FV/FM measurements obtained with the FMS-2 fluorometer in field were 267 found to be strongly correlated with those of detached needles at room 268 temperature (r=0.94, P<0.0001) (Fig. S3), suggesting that room temperature 269 measurements did not introduce any major bias into the observed seasonal 270 patterns in spectral ChlF. 11 271 Estimation of foliar pigment content 272 During each sampling point, five pairs of needles were randomly detached per 273 each tree using the pre-selected top branches and needle cohort. Needles were 274 sampled into cryotubes, immediately frozen at liquid nitrogen temperature using 275 a portable dewar (CX-100, Taylor Wharton International LLC, Minnetonka, MN), 276 and subsequently stored at -80 °C until extraction. Pigment analysis were 277 conducted following Wellburn (1994) with dimethyl sulfoxide (DMSO; VWR 278 Chemicals, 23500.322) as solvent. Frozen samples (75-100 mg) were first 279 homogenized for 2 minutes at 30 Hz, using a bead mill (TissueLyser II Qiagen, 280 Germany), stainless steel beads (4 mm) and microtubules (2 mL). Subsequently, 281 1.8 mL of DMSO was added to the homogenate and resuspended again at 30 Hz 282 for 1 minute. Pigments were extracted in oven at 40 °C for 4 hours. Extracts were 283 then centrifuged at 25000g for 5 minutes. Light absorption was measured at 284 649.1 nm, 665.1 nm and 480.0 nm, with a spectrophotometer (Shimadzu UV- 285 2401 PC), and subsequently used for estimation of Chl a, Chl b and total 286 carotenoids (Wellburn 1994). 287 288 Measurements of leaf absorption 289 Needle PAR absorption (AT(PAR)) was measured 8 times during the study period 290 (N=5 biological replicates). We used the same spectrometer and LED light source 291 described above connected to a 6 inch diameter integrating sphere 292 (AdaptaSphere, LabSphere Inc, New Hampshire, UK), and applied the black 293 spray method (Olascoaga, Mac Arthur, Atherton & Porcar-Castell 2016) to 294 estimate AT(PAR). The black spray method was inspired by earlier within-sphere 295 measurements (Öquist, Hällgren & Brunes 1978; Idle & Proctor 1983) and 12 296 purposefully developed to measure absorption in leaves with complex geometry, 297 like needles, because it does not require mounting needles in the port of the 298 integrating sphere and therefore there are no gap effects. Instead, samples are 299 placed inside the sphere and hung from a white thread across the central plane 300 (see illustrations in Olascoaga et al. 2016 for details). Prior to measurements, 5- 301 6 needle pairs were separated and the resulting 10-12 needles sewed with a 302 white thread and spaced at least 1 cm. The method consisted of three separate 303 spectral measurements inside the sphere: 1) white thread alone (reference zero 304 absorption, IW), 2) white thread with needles (sample absorption, IS), and 3) white 305 thread with needles painted with a black spray of known absorption (Black sample, 306 IB). Conveniently, because the total surface area of the needles in steps 2) and 307 3) can be assumed to remain constant, this parameter cancels out and there is 308 no need to estimate it. Total PAR absorption can then be computed by combining 309 these three measurements with the known absorption of the black spray (ABLACK), 310 as: (𝐼W −𝐼S ) 𝐼B 𝐴BLACK 311 𝐴T = (𝐼W −𝐼B ) 𝐼S Eqn. 4 312 The mean value of AT between 400 and 700nm was here used as an estimate of 313 AT(PAR). 314 315 Room temperature measurements of leaf spectral ChlF 316 Leaf spectral ChlF (Fλ, in mW m-2 sr-1 nm-1) was measured at room temperature 317 using a FluoWat Clip (Image Processing Laboratory, University of Valencia, 318 Spain) (Van Wittenberghe, Alonso, Verrelst, Moreno & Samson 2015) coupled 319 with a powerful white LED (MJ-858, Magicshine, UK) and a radiometrically 13 320 calibrated visible-near-infrared spectrometer (FieldSpec, ASD-Panalytical, 321 Boulder, CO). The spectrometer covers the spectral range between 325 and 1075 322 nm at a sampling interval of 1 nm and with a FWHM of 3.5 nm. The FluoWat clip 323 has an optical window for input illumination (incidence angle of 45°) and a small 324 aperture (at nadir view) to connect the optical fiber. The fiber field of view is 25° 325 and the distance to sample is 1 cm, yielding an approximately circular target area 326 of 0.22 cm radius. To obtain spectral ChlF, a 650 nm short-pass filter (Edmund 327 Optics Ltd, UK, OD=4) was used to exclude 99.99% of radiation above 650 nm. 328 The LED source supplies a PAR of c. 1200 µmol at the leaf surface as estimated 329 with a PAR sensor (Licor LI150-A, Li-Cor Inc., Lincoln, NE, USA). 330 Needles were carefully arranged alongside each other to minimize the gap 331 fraction and fixed with transparent tape to conform a needle mat (see Fig. 1). 332 Rajewicz, Atherton, Alonso & Porcar-Castell (2019) recently compared the ChlF 333 spectra of needles with different arrangements, and concluded that despite the 334 fact that needle mats could slightly enhance reabsorption, they provided a higher 335 replicability and reproducibility and were therefore a good solution to track 336 temporal changes in spectral properties. 337 Needle mats were dark adapted at 10 °C for 1 hour and kept in the dark until 338 measurements started in the FluoWat clip. After conducting a dark current 339 measurement, the needle mat was placed in the FluoWat clip, recording was 340 started, and an opaque aluminium foil placed between the light source and the 341 FluoWat clip rapidly removed. Variations of spectral ChlF were recorded during 342 two minutes at an integration time of 136 ms. The last ten spectra were averaged 343 and used to estimate steady-state ChlF between 650 nm and 850 nm (Fλ). 344 Averaged spectra were further smoothed using a Savitzky-Golay filter (order=2, 14 345 averaging interval=15) and cut to a range of 660-780 nm where the signal was 346 strongest. The resulting fluorescence spectra was used to estimate red (F690) and 347 far-red ChlF (F740) and used to calculate an integrated steady-state ChlF between 780 348 660 and 780 nm ( ∫660 𝐹 𝑑𝜆 ) for correspondence with the broadband PAM 349 fluorescence parameter, ΦF. 350 351 Statistical analyses 352 Mean values of four (gas exchange and Monitoring PAM fluorescence) or five (all 353 other data) trees were used to conduct statistical analysis. Linear regression 354 models and simple Pearson’s correlation analyses were used to assess 355 relationships between variables. All the analyses were conducted with R version 356 3.2.2 (R Core Development Team, 2015) and MATLAB version R2014a 357 (MathWorks Inc. 2014). 358 Principal Component Analysis (PCA) was used to quantitatively assess the 359 relative role of different components of variation in the spectral ChlF dataset. In 360 particular, we wanted to quantify how much of the seasonal variation in spectral 361 ChlF during the spring recovery of photosynthesis was related to principal 362 components associated with changes in the intensity and how much was related 363 to principal components associated with changes in the shape. The R function 364 ‘prcomp’ in the default mode ('prcomp' arguments; scale = FALSE and center = 365 TRUE) was used to run a PCA on time series of the ChlF data-set. The analysis 366 was focused on the 660-780 nm emission region to reduce impact of low signal- 367 to-noise at wavelength limits. 368 15 369 Results 370 Seasonal changes in micrometeorological variables and photosynthetic 371 parameters 372 Both temperature and PAR followed the typical annual pattern in boreal regions 373 with PAR increasing faster and earlier than temperature (Fig. 2a). Daily mean 374 temperatures in February (-1.7 °C) and March (0.6 °C) were higher by 6 and 4 °C, 375 respectively, compared to the average for the period 1981-2010 (Pirinen et al. 376 2012). Fortunately, a cold spell took place on March 22nd with a daily mean 377 temperature of -5.7 °C. On April 16th, a second cold spell took place with a mean 378 temperature of 0.7 °C. These cold spells served to assess the responses of ChlF 379 and photosynthetic parameters to a sudden decrease in temperature and were 380 highlighted in figures reporting the time series. 381 Noon mean CO2 fluxes (Fig. 2b) registered a clear seasonal pattern gradually 382 increasing from April to June. Both α (a measure of maximum LUE under low light) 383 and β (a measure of maximum photosynthetic rate) displayed similar seasonal 384 patterns (Fig. 2c), gradually increasing from winter to summer. The effect of the 385 cold spells on α and β parameters could not be distinguished. 386 Changes in the FV/FM and ΦF (Fig. 2d) measured in the field with the MONI-PAM 387 system also tracked the development of the spring recovery of photosynthesis, 388 rapidly decreasing in response to the two cold spells and gradually recovering 389 from winter to summer. Over the full spring recovery, ΦF varied approximately by 390 a factor of 2 from its minimum in early April to its maximum in June and July. 391 Variations in FV/FM were further decomposed into PQS and NPQS components 392 (Fig. 2e,f). NPQS presented larger seasonal variation decreasing from about 5 at 393 the beginning of the study period to zero during summer. In contrast, PQ S 16 394 increased from values of about 3 at the beginning of the study period to values of 395 4.5 during summer. Both NPQS and PQS consistently responded to both cold 396 spells, increasing and decreasing, respectively. 397 398 Seasonal changes in pigment content and leaf absorption 399 No obvious seasonal patterns were observed in total Chl levels (Fig. 3a) although 400 a slightly decreasing trend occurred during spring which reversed towards 401 summer. This result was consistent with the time-series of total leaf PAR 402 absorption (AT(PAR), Fig. S2) which did not display any seasonal pattern either 403 with values ranging from 0.81 to 0.85. In contrast, seasonal changes were 404 observed in Chl b (Fig. 3c), Chl a/b (Fig. 3d), carotenoid content (Fig. 3e) and 405 Car/Chl ratios (Fig. 3f). During early spring, Chl a/b ratios were lower and 406 carotenoid and Car/Chl ratios higher than during summer. No clear effects of the 407 cold spells were seen in the pigments, except for Chl a/b which displayed higher 408 levels after the second cold spell and Car/Chl which had a tendency to increase 409 after the cold spells. 410 411 Seasonal variation in ChlF spectra 412 The ChlF spectra of pine needles measured at room temperature and at standard 413 illumination (Fλ, Fig. 4a-c) was used here as a proxy of variations in the spectrally 414 resolved fluorescence yield. Fλ experienced strong seasonal changes during the 415 study period, mainly in terms of intensity but also in shape. For example, the red 416 peak near 690nm was nearly absent during winter (Fig. 4a,b), but gradually 417 reappeared towards summer (Fig. 4c). Both red (F690) and far-red (F740) ChlF (Fig. 17 418 4d) increased from early spring to summer with a larger change in F740. Integrated 780 419 ChlF (∫660 𝐹 𝑑𝜆, Fig. 4e), used here for comparison with the spectrally averaged 420 ΦF, presented a similar seasonal pattern with the rest of photosynthetic 780 421 parameters. Both ΦF (Fig. 2d) and ∫660 𝐹 𝑑𝜆 (Fig. 4e) varied seasonally by a 422 factor of 2. The F690/F740 ratio drastically decreased from 0.5 to 0.4 in response 423 to the first cold spell, and did not reverse to 0.5 until several weeks later (Fig. 4f). 424 Later on, the ratio decreased again to about 0.45 by the end of the study period, 425 coinciding with the increase in total Chl (Fig. 3a) and Fλ (Fig. 4d). 426 427 Principal component analysis 428 Three principal components explained 99.8% of the variation in ChlF spectra 429 during the study period (Fig. 5). PC1 explained 98.8% of variance of ChlF spectra 430 across the spring recovery (Fig. 5a) and exhibited a clear and similar seasonal 431 pattern (Fig. 5b) to that of red, far-red, and integrated ChlF (Fig. 4d,e). The 432 spectral weights and intensity of PC1 reflected the shape and seasonal variation 433 in ФF. In contrast, PC2 (Fig. 5c) explained only 0.8% of variance but its spectral 434 weights presented a strong and wavelength dependent shape reflecting 435 differences in red and far-red ChlF. PC2 potentially indicates seasonal changes 436 in ChlF reabsorption (i.e. fesc (λ) in Eqn.1), changes in energy redistribution 437 between PSII and PSI, or changes in the emission spectra at the photosystem 438 level. Remarkably, a drastic increase of PC2 (Fig. 5d) was detected upon the first 439 cold spell (solid and vertical black line), and later on during summer when foliar 440 Chl content tended to increase (Fig. 3a). Overall, the seasonal pattern of PC2 441 closely resembled the mirror image of F690/F740 (Fig. 4f). PC3 explained only 0.2% 18 442 of variance (Fig. 5e) and displayed no clear seasonal trend other than a sharp 443 decrease at the end of April (Fig. 5f). 444 445 Seasonal correlations between photosynthetic parameters and spectral ChlF 446 Measurements of red, far-red and integral ChlF under standardized conditions 447 were strongly correlated with field photosynthetic parameters α (r: 0.81-0.83), β 448 (r: 0.84-0.86) and FV/FM (r: 0.87-0.89) (P<0.001 for all; Fig. 6 and Fig. S4). In 449 contrast, no significant relationships between F690/F740 and photosynthetic 450 parameters were found. Red (F690) and far-red (F740) ChlF were strongly and 451 negatively correlated also with leaf carotenoid content (r: -0.9 and -0.87, 452 respectively) and NPQS (r: -0.86 and -0.83, respectively), and positively related 453 with ΦF (r: 0.73 and 0.71, respectively) and PQS (r: 0.74 and 0.71, respectively) 454 (P<0.001 for all; Fig. 6 and Fig. S4). In fact, NPQS and leaf carotenoid content 455 were the two factors that displayed the strongest (negative) correlation with the 456 spring recovery of photosynthesis in terms of α, β, and FV/FM. No significant 457 relationship was observed between red or far-red ChlF and total Chl or AT(PAR) 458 (Fig. S4). 459 As for the principal components, PC1 was correlated strongly with FV/FM (r = 0.87), 460 and negatively correlated with carotenoid, Car/Chl, and NPQ S (r: -0.87 to -0.83), 461 and had a significant but slightly lower relationship with ΦF and PQS (r = 0.72) 462 (P<0.001 for all; Fig. 6 and Fig. S4). PC2 was highly and negatively correlated 463 with F690/F740 (r = -0.92; P<0.001) but was not significantly correlated with AT(PAR) 464 (r = 0.51) and total Chl (r = 0.37). Finally, PC3 was only marginally correlated with 465 Chl a/b (r = 0.45; P = 0.06). 19 466 We also conducted a systematic assessment of the seasonal correlations 467 between different ChlF emission wavelengths and photosynthetic parameters 468 (Fig. 7). The results indicated that both red and far-red ChlF wavelengths were 469 strongly correlated with the spring recovery of photosynthesis (P<0.001 for all), 470 represented here by the photosynthetic parameters α and β (Fig. 7a), with red 471 wavelengths displaying stronger correlations than far-red wavelengths. 472 Correlations were also stronger for β, compared to α across the whole spectral 473 range, which is not surprising since our spectral measurements were conducted 474 at high light. The strength of the correlation decreased between the two ChlF 475 emission peak, with highest RMSE (data not shown). As expected, a wavelength 476 dependency in the slope of α (β) vs. Fλ was also observed (Fig. 7b), further 477 emphasizing that the ratio LUE/ФF(λ) (as in Eqn. 3) was dependent on ChlF 478 emission wavelengths. Positive intercepts of Fλ vs. α (β) (Fig. 7c; with ranges of 479 0.15 to 0.95 mW m-2 sr-1 nm-1) were obtained for both parameters and across 480 wavelengths, indicating positive ChlF emission when photosynthetic gas 481 exchange approaches zero. 482 Additionally, all the main ChlF emission wavelengths from 680 to 770 nm were 483 well correlated with ФF (P<0.001 for all), and presented slightly stronger 484 relationships in the red region than far-red (Fig. S5a). This was consistent with 485 better correlations of ChlF with photosynthetic parameters in the red than in the 780 486 far-red wavelengths (Fig. 7a). Further, when comparing Fλ to ∫660 𝐹 𝑑𝜆 , red 487 wavelengths presented slightly lower correlations (Fig. S5b), reflecting that red 488 ChlF fluorescence wavelengths account for a source of independent information 489 not conveyed by broadband ChlF. 490 20 491 Discussion 492 We followed the spectral response of ChlF during the spring recovery of 493 photosynthesis. We found an across wavelength increase in ChlF ‘level’ well 494 correlated with photosynthetic parameters, and additional subtler changes in 495 spectral shape over time. Critically, photosystems in overwintering evergreens 496 undergo major structural and biochemical adjustments during the non- 497 photosynthetic season (Adams & Demmig-Adams 1994; Ottander et al. 1995; 498 Gilmore & Ball 2000; Ensminger et al. 2004; Verhoeven 2014), which were 499 reflected in the variation of the measured ChlF spectra. 500 501 Seasonal variation in leaf-level ChlF spectra and its controls 502 We used PCA analysis to separate and quantify the relative roles of different 503 components of variation in the seasonal ChlF dynamics of pine needles. We 504 found that 98.8% of the seasonal variation in ChlF spectra of needles was 505 explained by changes in ChlF level (PC1) with only a marginal 1% of variation 506 associated with additional changes in shape (with PC2 explaining 0.8%). Factors 507 contributing to these components of variation are discussed next. 508 Lack of correlation between F690, F740 (and PC1), and total Chl content or AT(PAR) 509 (Figs. 6 and S4) evidenced the minor role of PAR absorption in driving the 510 seasonality of ChlF during the spring recovery. In the present study Chl content 511 remained relatively stable (Fig. 3a). This is at odds with previous work where 512 significant seasonal changes in Chl content have been observed in Scots pine 513 needles (Ottander et al. 1995; Ensminger et al. 2004, Porcar-Castell et al. 2012), 514 and other evergreen species (Wong & Gamon 2015). It is unclear why we 21 515 observed such differences in pigments; but considering the alternative scenario 516 where Chl did change during the season, and as the bulk of the spring recovery 517 of ChlF in Scots pine needles has been shown to precede the summer increase 518 in foliar Chl content by about two weeks (Porcar-Castell et al. 2008a), changes in 519 leaf PAR absorption (A, Eqn.1) will likely remain of lesser importance in 520 controlling the seasonality in leaf-level SIF(λ) in evergreen conifers. We also did 521 not find evidence of an inverse and persistent relationship between red/far-red 522 ratio and foliar Chl content across the study period (Figs. 6 and S4), as would 523 follow from the general Chl reabsorption theory (Fig. 4A in Gitelson, Buschmann 524 & Lichtenthaler 1999; Fig. 4 in Buschmann 2007). This result demonstrates that, 525 in addition to foliar Chl content, other factors also influence the shape of the ChlF 526 spectra. 527 Seasonal changes in both F690 and F740 (and PC1) were strongly related to 528 seasonal variation in foliar carotenoid content, NPQs, and to a lesser extent PQs 529 (Fig. 6). The mechanisms that drive NPQs in overwintering evergreens 530 (understood as the sustained enhancement in the capacity for thermal dissipation 531 of excitation energy), remain under intense investigation (Demmig-Adams & 532 Adams 2006; Verhoeven 2014; Ruban 2016; Malnoë 2018). NPQs in evergreens 533 has been associated to: the accumulation of carotenoids (especially zeaxanthin) 534 relative to Chl (Adams & Demmig-Adams 1994; Ensminger et al. 2004; Zarter et 535 al. 2006a; Porcar-Castell et al. 2012), the accumulation of damaged or non- 536 functional PSII reaction centers coupled to a reduction in the PSII core D1 protein 537 (Ottander et al. 1995; Ensminger et al. 2004; Ebbert, Adams, Mattoo, Sokolenko 538 & Demmig-Adams 2005), the presence and accumulation of early-light induced 539 proteins (Elips) with or without concomitant changes in minor antenna PsbS 22 540 proteins (Ensminger et al. 2004; Ebbert et al. 2005; Zarter et al. 2006b; 541 Verhoeven 2014), the aggregation of light-harvesting complexes of PSII (LHCII) 542 (Horton et al. 1991; Ottander et al. 1995; Busch, Hüner & Ensminger 2007; Ruban 543 2018), and the unstacking of the thylakoid membrane (Demmig-Adams et al. 544 2015). 545 Reversible forms of NPQ, which operate in time scales of seconds to minutes, 546 (e.g. energy-dependent or zeaxanthin-dependent quenching qE and qZ, 547 respectively) (Verhoeven 2014; Malnoë 2018), are known to predominantly 548 quench ChlF associated with PSII units with no apparent effect on PSI ChlF 549 (Genty et al. 1990; Franck et al. 2002). As a result, because the ChlF contribution 550 from PSI is much larger around the far-red peak (Genty et al. 1990; Pfündel 1998; 551 Franck et al. 2002), an increase in qE or qZ results in a reduction in the red/far- 552 red ChlF ratio (Agati, Mazzinghi, Fusi & Ambrosini 1995; Agati et al. 2000). If the 553 PSI ChlF contribution would be also insensitive to NPQs, we would have 554 expected a similar decrease in F690/F740 in response to the seasonal accumulation 555 of NPQs. This scenario was not fully supported by our observations because the 556 extremes of variation in F690/F740 ratio (observed during March, Fig. 4f) took place 557 under comparable Chl and NPQs levels (Figs. 2f and 3a), suggesting that PSI 558 ChlF contribution is dynamic at the seasonal scale. 559 Seasonal changes in the ChlF ratio are clearly driven by multiple controls. For 560 example, F690/F740 decreased from 0.5 to 0.4 in response to the first cold spell, 561 along with an increase in NPQs (Fig. 2f) and in PC2 (Fig. 5d) which occurred 562 under rather stable Chl content (Fig. 3a). This phenomena could denote a 563 structural reorganization at the level of LHCII (increased reabsorption due to 564 aggregation), or thylakoid membrane unstacking (promoting energetic 23 565 connectivity between PSII and PSI units) accompanying the accumulation of 566 NPQs. In fact, similar decreases in red/far-red ChlF ratio have been reported in 567 response to reversible NPQ-induced oligomerization of LHCII complexes 568 (Miloslavina et al. 2008; Jahns & Holzwarth 2012). Later on, in mid-May and July 569 we registered a slight increase in foliar Chl content (Fig. 3a) accompanied by a 570 decrease in the F690/F740 ratio (Fig. 4f), and under rather stable NPQs (Fig. 2f) 571 corresponding to the previous characterized relationship between leaf Chl 572 content and the ChlF ratio (Gitelson et al. 1999). Overall, further experimental 573 and modelling studies are needed to clarify the mechanisms that drive the 574 dynamics in the ChlF spectra in overwintering evergreens. 575 576 Wavelength dependency between ChlF and photosynthesis 577 We found that all the main ChlF emission wavelengths (680-770 nm) were 578 correlated with and capable of tracking the photosynthetic spring recovery in 579 boreal evergreen foliage (Fig. 7a). This finding was consistent with the dominant 580 role of NPQs in controlling the seasonal variability in the ChlF level across 581 wavelengths (PC1; Fig. 5a,b). It is important to note however that the relative 582 contributions of the variation in spectral shape (1% for PC2+PC3) and ChlF level 583 (98.8% for PC1) to total seasonal variation in ChlF spectral properties could have 584 been very different in a different species or under a different set of environmental 585 conditions, encouraging further seasonal studies across other species and 586 biomes. 587 Slightly higher correlations (Fig. 7a) and slopes (Fig. 7b) were found between red 588 wavelengths of Fλ and α and β. Similar results have been reported at the canopy 589 scale based on model simulations (Verrelst et al. 2015; Liu et al. 2019) and 24 590 experimental data (Cheng et al. 2013; Campbell et al. 2019; Magney et al. 2019a). 591 At the leaf-level, the higher correlation in red wavelengths could point to 592 interferences with the dynamics or PSI ChlF in the far-red wavelengths. In fact, 593 the spectral shape of the intercept (Fig. 7c), which represents the background 594 ChlF spectra at zero photosynthesis, had a high resemblance to the typical 595 spectral shape of PSI ChlF, although the peak was slightly red-shifted (740 nm) 596 compared to earlier studies in non-downregulated leaves of barley (722 nm) 597 (Franck et al. 2002) or maize PSI particles (725-730 nm) (Croce, Dorra, 598 Holzwarth & Jennings 2000). Further work to characterize and identify the drivers 599 of this background ChlF signal, which may have important implications for 600 interpretation of SIF data over boreal evergreen regions, is needed. 601 602 Implications at the larger scale 603 SIF is a promising methodology for the estimation of GPP dynamics in terrestrial 604 ecosystems either using statistical methods based on Eqn. 3 (Guanter et al. 2014; 605 Sun et al. 2017; Li et al. 2018), or by assimilation into the photosynthetic modules 606 of land surface models (Lee et al. 2015; Thum et al. 2017; Macbean et al. 2018). 607 SIF could be particularly useful for evergreen ecosystems by capturing part of the 608 temporal dynamics of GPP that remains hidden to traditional reflectance-based 609 vegetation indices due to low variation in greenness (Walther et al. 2016; Smith 610 et al. 2018; Zuromski et al. 2018; Magney et al. 2019a; Nichol et al. 2019). 611 Overall, our results indicate that current Fraunhofer- and oxygen-based methods 612 to retrieve SIF across different wavelengths in the red and near-infrared regions 613 (Meroni et al. 2009; Lu, Cheng, Li & Tang 2018) have similar intrinsic potential to 614 capture the leaf-level spring recovery of photosynthesis in boreal evergreen 25 615 forests. Importantly, when upscaling from the leaf to the canopy and landscape 616 level, the constancy assumption behind the LUE/ФF(λ) factor in Eqn. 3 will be 617 further complicated by spatial variation in physiological factors caused by within 618 canopy light and temperature gradients or species composition (Porcar-Castell 619 et al. 2014; Sun et al. 2017). Similarly, canopy-level fesc (Eqn. 3) will be also 620 affected by seasonal dynamics in canopy structure related to the phenology of 621 the multiple components in the ecosystem, including the understory (Majasalmi, 622 Stenberg & Rautiainen 2017; Liu et al. 2019). 623 624 Conclusions 625 We demonstrated that red and far-red ChlF emission wavelengths were able to 626 transmit the optical signature of the spring recovery of photosynthesis in boreal 627 evergreen needles. The dynamics of the spectral signature were mediated by the 628 complex and highly articulated process of sustained regulatory thermal 629 dissipation or NPQS, which effectively quenches all ChlF wavelengths. Although 630 changes in the shape of the ChlF spectra explained only a marginal proportion of 631 the observed seasonality, the information content embedded in these changes 632 could be highly informative. Specifically, the rapid decrease in the F690/F740 ratio 633 upon the first cold spell suggests that in addition to Chl content, F690/F740 can also 634 convey information on the structural organization in the thylakoid membrane in 635 overwintering evergreens, which could help to better constrain the assimilation of 636 SIF data into models of photosynthesis (Macbean et al. 2018; Raczka et al. 2019). 637 Further investigations across species and scales will be required to fully 638 characterize the information potentially embedded in the spectral dynamics of SIF, 639 in support of multispectral SIF retrievals from towers, drones, airplanes as well 26 640 as satellite missions such as current TROPOspheric Monitoring Instrument, 641 TROPOMI, onboard Sentinel-5 (Guanter et al. 2015; Köhler et al. 2018) or future 642 Fluorescence Explorer mission, FLEX (Drusch et al. 2017). 643 644 Acknowledgements: We acknowledge the financial support from the Academy 645 of Finland (288039, 293443, 319211, 272041), the COST Action 646 ES1309/OPTIMISE, the European Research Council Synergy grant SyG-2013- 647 610028 IMBALANCE-P, the Spanish Government project CGL2016-79835-P, the 648 Catalan Government project SGR 2014-274, and the Funds from the University 649 of Helsinki (Grant 490116). HR was supported by EU LIFE 12 ENV/FI/000409 650 Monimet. CZ gratefully acknowledges the support from the China Scholarship 651 Council. 652 653 Conflict of interest: The authors have no conflicts of interest to declare. 654 655 Author Contribution: Albert Porcar-Castell, Jon Atherton, Iolanda Filella, Josep 656 Peñuelas and Jaana Bäck conceived and designed the study; Chao Zhang and 657 Albert Porcar-Castell carried out the optical measurements; Chao Zhang and Jon 658 Atherton conducted the majority of the data analysis; Pasi Kolari estimated the 659 photosynthetic parameters; Juho Aalto carried out and processed the field gas 660 exchange data at SMEAR-II station; Hanna Ruhanen conducted the pigments 661 analysis; Chao Zhang and Albert Porcar-Castell wrote the paper with 662 contributions from all authors. 663 27 664 References 665 Aalto J., Kolari P., Hari P., Kerminen V.M., Schiestl-Aalto P., Aaltonen H., … 666 Bäck J. (2014) New foliage growth is a significant, unaccounted source for 667 volatiles in boreal evergreen forests. Biogeosciences 11, 1331–1344. 668 Adams W.W. & Demmig-Adams B. (1994) Carotenoid composition and down 669 regulation of photosystem II in three conifer species during the winter. 670 Physiologia Plantarum 92, 451–458. 671 Agati G., Cerovic Z.G. & Moya I. (2000) The effect of decreasing temperature 672 up to chilling values on the in vivo F685/F735 chlorophyll fluorescence ratio 673 in Phaseolus vulgaris and Pisum sativum: the role of the photosystem I 674 contribution to the 735 nm fluorescence band. Photochemistry and 675 Photobiology 72, 75–84. 676 Agati G., Mazzinghi P., Fusi F. & Ambrosini I. (1995) The F685/F730 chlorophyll 677 fluorescence ratio as a tool in plant physiology: response to physiological 678 and environmental factors. Journal of Plant Physiology 145, 228–238. 679 Alonso L., Gómez-chova L., Vila-francés J., Amorós-lópez J., Guanter L., Calpe 680 J. & Moreno J. (2007) Sensitivity analysis of the Fraunhofer Line 681 Discrimination method for the measurement of chlorophyll fluorescence 682 using a field spectroradiometer. IEEE, 3756–3759. 683 Baker N.R. (2008) Chlorophyll fluorescence: a probe of photosynthesis in vivo. 684 Annual review of plant biology 59, 89–113. 685 Barber J., Malkin S. & Telfer A. (1989) The origin of chlorophyll fluorescence in 686 vivo and its quenching by the photosystem II reaction centre. Philosophical 687 Transactions of the Royal Society B: Biological Sciences 323, 227–239. 28 688 Busch F., Hüner N.P.A. & Ensminger I. (2007) Increased air temperature during 689 simulated autumn conditions does not increase photosynthetic carbon gain 690 but affects the dissipation of excess energy in seedlings of the evergreen 691 conifer jack pine. Plant Physiology 143, 1242–1251. 692 Buschmann C. (2007) Variability and application of the chlorophyll fluorescence 693 emission ratio red/far-red of leaves. Photosynthesis Research 92, 261–271. 694 Campbell P.K., Huemmrich K.F., Middleton E.M., Ward L.A., Julitta T., Daughtry 695 C.S., … Kustas W.P. (2019) Diurnal and seasonal variations in chlorophyll 696 fluorescence associated with photosynthesis at leaf and canopy scales. 697 Remote Sensing 11, 1–36. 698 Cheng Y., Middleton E.M., Zhang Q., Huemmrich K.F., Campbell P.K.E., Corp 699 L.A., … Daughtry C.S. (2013) Integrating solar induced fluorescence and 700 the photochemical reflectance index for estimating gross primary 701 production in a cornfield. Remote Sensing 5, 6857–6879. 702 Croce R., Dorra D., Holzwarth A.R. & Jennings R.C. (2000) Fluorescence decay 703 and spectral evolution in intact photosystem I of higher plants. Biochemistry 704 39, 6341–6348. 705 Damm A., Guanter L., Paul-limoges E., Tol C. Van Der, Hueni A., Buchmann 706 N., … Schaepman M.E. (2015) Far-red sun-induced chlorophyll 707 fluorescence shows ecosystem-specific relationships to gross primary 708 production: An assessment based on observational and modeling 709 approaches. Remote Sensing of Environment 166, 91–105. 710 Demmig-Adams B. & Adams W.W. (2006) Photoprotection in an ecological 711 context: the remarkable complexity of thermal energy dissipation. New 29 712 Phytologist 172, 11–21. 713 Demmig-Adams B., Muller O., Stewart J.J., Cohu C.M. & Adams W.W. (2015) 714 Chloroplast thylakoid structure in evergreen leaves employing strong 715 thermal energy dissipation. Journal of Photochemistry and Photobiology B: 716 Biology 152, 357–366. 717 Drusch M., Moreno J., Del Bello U., Franco R., Goulas Y., Huth A., … Verhoef 718 W. (2017) The FLuorescence EXplorer Mission Concept-ESA’s Earth 719 Explorer 8. IEEE Transactions on Geoscience and Remote Sensing 55, 720 1273–1284. 721 Ebbert V., Adams W.W., Mattoo A.K., Sokolenko A. & Demmig-Adams B. 722 (2005) Up-regulation of a photosystem II core protein phosphatase inhibitor 723 and sustained D1 phosphorylation in zeaxanthin-retaining, photoinhibited 724 needles of overwintering Douglas fir. Plant, Cell and Environment 28, 232– 725 240. 726 Ensminger I., Sveshnikov D., Campbell D.A.., Funk C., Jansson S., Lloyd J., … 727 Öquist G. (2004) Intermittent low temperatures constrain spring recovery of 728 photosynthesis in boreal Scots pine forests. Global Change Biology 10, 729 995–1008. 730 Franck F., Juneau P. & Popovic R. (2002) Resolution of the Photosystem I and 731 Photosystem II contributions to chlorophyll fluorescence of intact leaves at 732 room temperature. Biochimica et Biophysica Acta - Bioenergetics 1556, 733 239–246. 734 Frankenberg C. & Berry J. (2018) Solar induced chlorophyll fluorescence : 735 origins , relation to photosynthesis and retrieval. In Reference Module in 30 736 Earth Systems and Environmental Sciences: Comprehensive Remote 737 Sensing, Elsevier, Oxford. pp. 143–162. Elsevier. 738 Frankenberg C., Fisher J.B., Worden J., Badgley G., Saatchi S.S., Lee J.E., … 739 Yokota T. (2011) New global observations of the terrestrial carbon cycle 740 from GOSAT: Patterns of plant fluorescence with gross primary 741 productivity. Geophysical Research Letters 38, 1–6. 742 Frankenberg C., O’Dell C., Berry J., Guanter L., Joiner J., Köhler P., … Taylor 743 T.E. (2014) Prospects for chlorophyll fluorescence remote sensing from the 744 Orbiting Carbon Observatory-2. Remote Sensing of Environment 147, 1– 745 12. 746 Genty B., Wonders J. & Baker N.R. (1990) Non-photochemical quenching of Fo 747 in leaves is emission wavelength dependent: consequences for quenching 748 analysis and its interpretation. Photosynthesis Research 26, 133–139. 749 Gilmore A.M. & Ball M.C. (2000) Protection and storage of chlorophyll in 750 overwintering evergreens. Proceedings of the National Academy of 751 Sciences 97, 11098–11101. 752 Gitelson A.A., Buschmann C. & Lichtenthaler H.K. (1999) The chlorophyll 753 fluorescence ratio F735 / F700 as an accurate measure of the chlorophyll 754 content in plants. Remote Sensing of Environment 69, 296–302. 755 Govindjee G. (1995) Sixty-three years since Kautsky: chlorophyll a 756 fluorescence. Australian Journal of Plant Physiology 22, 131–160. 757 Guanter L., Aben I., Tol P., Krijger J.M., Hollstein A., Köhler P., … Landgraf J. 758 (2015) Potential of the TROPOspheric Monitoring Instrument (TROPOMI) 759 onboard the Sentinel-5 Precursor for the monitoring of terrestrial chlorophyll 31 760 fluorescence. Atmospheric Measurement Techniques 8, 1337–1352. 761 Guanter L., Zhang Y., Jung M., Joiner J., Voigt M., Berry J.A., … Moran M.S. 762 (2014) Global and time-resolved monitoring of crop photosynthesis with 763 chlorophyll fluorescence. Proceedings of the National Academy of 764 Sciences, 1–7. 765 Hari P. & Kulmala M. (2005) Station for measuring ecosystem-atmosphere 766 relations (SMEAR II). Boreal Environment Research 10, 351–322. 767 Hari P., Mäkelä A., Korpilahti E. & Holmberg M. (1986) Optimal control of gas 768 exchange. Tree Physiology 2, 169–175. 769 Horton P., Ruban A. V., Rees D., Pascal A.A., Noctor G. & Young A.J. (1991) 770 Control of the light-harvesting function of chloroplast membranes by 771 aggregation of the LHCII chlorophyll-protein complex. FEBS Letters 292, 772 1–4. 773 Huang W., Yang Y.J., Hu H. & Zhang S.B. (2016) Seasonal variations in 774 photosystem I compared with photosystem II of three alpine evergreen 775 broad-leaf tree species. Journal of Photochemistry and Photobiology B: 776 Biology 165, 71–79. 777 Idle D.B. & Proctor C.W. (1983) An integrating sphere leaf chamber. Plant, Cell 778 and Environment 6, 437–439. 779 IPCC-SR15 (2018) Special report IPCC. “Global Warming of 1.5°C.” 780 Jahns P. & Holzwarth A.R. (2012) The role of the xanthophyll cycle and of lutein 781 in photoprotection of photosystem II. Biochimica et Biophysica Acta - 782 Bioenergetics 1817, 182–193. 32 783 Joiner J., Yoshida Y., Guanter L. & Middleton E.M. (2016) New methods for the 784 retrieval of chlorophyll red fluorescence from hyperspectral satellite 785 instruments: simulations and application to GOME-2 and SCIAMACHY. 786 Atmospheric Measurement Techniques 9, 3939–3967. 787 Kitajima M. & Butler W.L. (1975) Quenching of chlorophyll fluorescence and 788 primary photochemistry in chloroplasts by dibromothymoquinone. 789 Biochimica et Biophysica Acta 376, 105–115. 790 Köhler P., Frankenberg C., Magney T.S., Guanter L., Joiner J. & Landgraf J. 791 (2018) Global retrievals of solar-induced chlorophyll fluorescence with 792 TROPOMI: first results and intersensor comparison to OCO-2. Geophysical 793 Research Letters 45, 10,456-10,463. 794 Kolari P., Bäck J., Taipale R., Ruuskanen T.M., Kajos M.K., Rinne J., … Hari P. 795 (2012) Evaluation of accuracy in measurements of VOC emissions with 796 dynamic chamber system. Atmospheric Environment 62, 344–351. 797 Kolari P., Chan T., Porcar-Castell A., Bäck J., Nikinmaa E. & Juurola E. (2014) 798 Field and controlled environment measurements show strong seasonal 799 acclimation in photosynthesis and respiration potential in boreal Scots pine. 800 Frontiers in Plant Science 5, 717. 801 Kolari P., Lappalainen H.K., Hänninen H. & Hari P. (2007) Relationship between 802 temperature and the seasonal course of photosynthesis in Scots pine at 803 northern timberline and in southern boreal zone. Tellus B: Chemical and 804 Physical Meteorology 59B, 542–552. 805 Lee J.-E., Berry J.A., van der Tol C., Yang X., Guanter L., Damm A., … 806 Frankenberg C. (2015) Simulations of chlorophyll fluorescence 33 807 incorporated into the Community Land Model version 4. Global Change 808 Biology 21, 3469–3477. 809 Li X., Xiao J., He B., Arain M.A., Beringer J., Desai A.R., … Varlagin A. (2018) 810 Solar-induced chlorophyll fluorescence is strongly correlated with terrestrial 811 photosynthesis for a wide variety of biomes: First global analysis based on 812 OCO-2 and flux tower observations. Global Change Biology, 1–19. 813 Liu W., Atherton J., Mõttus M., Gastellu-Etchegorry J.P., Malenovský Z., 814 Raumonen P., … Porcar-Castell A. (2019) Simulating solar-induced 815 chlorophyll fluorescence in a boreal forest stand reconstructed from 816 terrestrial laser scanning measurements. Remote Sensing of Environment 817 in press. 818 Lu X., Cheng X., Li X. & Tang J. (2018) Opportunities and challenges of 819 applications of satellite-derived sun-induced fluorescence at relatively high 820 spatial resolution. Science of the Total Environment 619–620, 649–653. 821 Macbean N., Maignan F., Bacour C., Lewis P., Guanter L., Köhler P., … Disney 822 M. (2018) Strong constraint on modelled global carbon uptake using solar- 823 induced chlorophyll fluorescence data. Scientific reports, 1–12. 824 Magney T.S., Bowling D.R., Logan B., Grossmann K., Stutz J., Blanken P., … 825 Frankenberg C. (2019a) Mechanistic evidence for tracking the seasonality 826 of photosynthesis with solar-induced fluorescence. Proceedings of the 827 National Academy of Sciences 116, 11640–11645. 828 Magney T.S., Frankenberg C., Kӧhler P., North G., Davis T.S., Dold C., … 829 Porcar-Castell A. (2019b) Disentangling changes in the spectral shape of 830 chlorophyll fluorescence: Implications for remote sensing of 34 831 photosynthesis. Journal of Geophysical Research: Biogeosciences 124, 1– 832 17. 833 Majasalmi T., Stenberg P. & Rautiainen M. (2017) Comparison of ground and 834 satellite-based methods for estimating stand-level fPAR in a boreal forest. 835 Agricultural and Forest Meteorology 232, 422–432. 836 Malnoë A. (2018) Photoinhibition or photoprotection of photosynthesis? Update 837 on the (newly termed) sustained quenching component qH. Environmental 838 and Experimental Botany 154, 123–133. 839 Meroni M., Rossini M., Guanter L., Alonso L., Rascher U., Colombo R. & 840 Moreno J. (2009) Remote sensing of solar-induced chlorophyll 841 fluorescence: review of methods and applications. Remote Sensing of 842 Environment 113, 2037–2051. 843 Migliavacca M., Perez-Priego O., Rossini M., El-Madany T.S., Moreno G., van 844 der Tol C., … Reichstein M. (2017) Plant functional traits and canopy 845 structure control the relationship between photosynthetic CO2 uptake and 846 far-red sun-induced fluorescence in a Mediterranean grassland under 847 different nutrient availability. New Phytologist 214, 1078–1091. 848 Miloslavina Y., Wehner A., Lambrev P.H., Wientjes E., Reus M., Garab G., … 849 Holzwarth A.R. (2008) Far-red fluorescence: A direct spectroscopic marker 850 for LHCII oligomer formation in non-photochemical quenching. FEBS 851 Letters 582, 3625–3631. 852 Monteith J.L. (1972) Solar radiation and productivity in tropical ecosystems. 853 Journal of Applied Ecology 9, 747–766. 854 Murata N., Takahashi S., Nishiyama Y. & Allakhverdiev S.I. (2007) 35 855 Photoinhibition of photosystem II under environmental stress. Biochimica et 856 Biophysica Acta - Bioenergetics 1767, 414–421. 857 Murchie E.H. & Lawson T. (2013) Chlorophyll fluorescence analysis: a guide to 858 good practice and understanding some new applications. Journal of 859 Experimental Botany 64, 3983–3998. 860 Nichol C., Drolet G., Porcar-Castell A., Wade T., Sabater N., Middleton E., … 861 Atherton J. (2019) Seasonal solar induced chlorophyll fluorescence and 862 photosynthesis in a boreal Scots pine canopy. Remote Sensing 11, 273. 863 Olascoaga B., Mac Arthur A., Atherton J. & Porcar-Castell A. (2016) A 864 comparison of methods to estimate photosynthetic light absorption in 865 leaves with contrasting morphology. Tree Physiology 36, 368–379. 866 Öquist G., Chow W.S. & Anderson J.M. (1992) Photoinhibition of 867 photosynthesis represents a mechanism for the long-term regulation of 868 photosystem II. Planta 186, 450–460. 869 Öquist G., Hällgren J.E. & Brunes L. (1978) An apparatus for measuring 870 photosynthetic quantum yields and quanta absorption spectra of intact 871 plants. Plant, Cell and Environment 1, 21–27. 872 Öquist G. & Huner N.P.A. (2003) Photosynthesis of overwintering evergreen 873 plants. Annual Review of Plant Biology 54, 329–355. 874 Ottander C., Campbell D. & Öquist G. (1995) Seasonal changes in photosystem 875 II organisation and pigment composition in Pinus sylvestris. Planta 197, 876 176–183. 877 Ottander C. & Öquist G. (1991) Recovery of photosynthesis in winter-stressed 878 Scots pine. Plant, Cell and Environment 14, 345–349. 36 879 Palombi L., Cecchi G., Lognoli D., Raimondi V., Toci G. & Agati G. (2011) A 880 retrieval algorithm to evaluate the photosystem I and photosystem II 881 spectral contributions to leaf chlorophyll fluorescence at physiological 882 temperatures. Photosynthesis research 108, 225–239. 883 Parazoo N.C., Arneth A., Pugh T.A.M., Smith B., Steiner N., Luus K., … Miller 884 C. (2018) Spring photosynthetic onset and net CO2 uptake in Alaska 885 triggered by landscape thawing. Global Change Biology, 3416–3435. 886 Parazoo N.C., Bowman K., Fisher J.B., Frankenberg C., Jones D.B.A., Cescatti 887 A., … Montagnani L. (2014) Terrestrial gross primary production inferred 888 from satellite fluorescence and vegetation models. Global Change Biology 889 20, 3103–3121. 890 Pfündel E. (1998) Estimating the contribution of photosystem I to leaf 891 chlorophyll fluorescence. Photosynthesis Research 56, 185–195. 892 Pfündel E.E., Klughammer C., Meister A. & Cerovic Z.G. (2013) Deriving 893 fluorometer-specific values of relative PSI fluorescence intensity from 894 quenching of F0 fluorescence in leaves of Arabidopsis thaliana and Zea 895 mays. Photosynthesis research 114, 189–206. 896 Pirinen P., Simola H., Aalto J., Kaukoranta J.-P., Karlsson P. & Ruuhela R. 897 (2012) Tilastoja suomen ilmastosta 1981-2010. (Climatological statistics of 898 Finland 1981–2010). 899 Porcar-Castell A. (2011) A high-resolution portrait of the annual dynamics of 900 photochemical and non-photochemical quenching in needles of Pinus 901 sylvestris. Physiologia Plantarum 143, 139–153. 902 Porcar-Castell A., Mac Arthur A., Rossini M., Eklundh L., Pacheco-Labrador J., 37 903 Anderson K., … Vescovo L. (2015) EUROSPEC: At the interface between 904 remote-sensing and ecosystem CO2 flux measurements in Europe. 905 Biogeosciences 12, 6103–6124. 906 Porcar-Castell A., Garcia-Plazaola J.I., Nichol C.J., Kolari P., Olascoaga B., 907 Kuusinen N., … Nikinmaa E. (2012) Physiology of the seasonal relationship 908 between the photochemical reflectance index and photosynthetic light use 909 efficiency. Oecologia 170, 313–323. 910 Porcar-Castell A., Juurola E., Ensminger I., Berninger F., Hari P. & Nikinmaa E. 911 (2008a) Seasonal acclimation of photosystem II in Pinus sylvestris. II. 912 Using the rate constants of sustained thermal energy dissipation and 913 photochemistry to study the effect of the light environment. Tree physiology 914 28, 1483–1491. 915 Porcar-Castell A., Pfündel E., Korhonen J.F.J. & Juurola E. (2008b) A new 916 monitoring PAM fluorometer ( MONI-PAM) to study the short- and long- 917 term acclimation of photosystem II in field conditions. Photosynthesis 918 research 96, 173–179. 919 Porcar-Castell A., Tyystjärvi E., Atherton J., van der Tol C., Flexas J., Pfündel 920 E.E., … Berry J.A. (2014) Linking chlorophyll a fluorescence to 921 photosynthesis for remote sensing applications: mechanisms and 922 challenges. Journal of Experimental Botany 65, 4065–4095. 923 Quéré C. Le, Andrew R.M., Friedlingstein P., Sitch S., Hauck J., Pongratz J., … 924 Adrie and B.Z. (2018) Global Carbon Budget 2018. Earth System Science 925 Data 10, 2141–2194. 926 Raczka B., Porcar‐Castell A., Magney T., Lee J.E., Köhler P., Frankenberg 38 927 C., … Bowling D.R. (2019) Sustained non‐photochemical quenching 928 shapes the seasonal pattern of solar‐induced fluorescence at a high‐ 929 elevation evergreen forest. Journal of Geophysical Research: 930 Biogeosciences 124. 931 Rajewicz P., Atherton J., Alonso L. & Porcar-Castell A. (2019) Leaf-level 932 spectral fluorescence measurements: comparing methodologies for 933 broadleaves and needles. Remote Sensing 11, 1–20. 934 Rascher U., Alonso L., Burkart A., Cilia C., Cogliati S., Colombo R., … Zemek 935 F. (2015) Sun-induced fluorescence – a new probe of photosynthesis: First 936 maps from the imaging spectrometer HyPlant. Global Change Biology 21, 937 4673–4684. 938 Romero J.M., Cordon G.B. & Lagorio M.G. (2018) Modeling re-absorption of 939 fluorescence from the leaf to the canopy level. Remote Sensing of 940 Environment 204, 138–146. 941 Ruban A. V. (2018) Light harvesting control in plants. FEBS Letters 592, 3030– 942 3039. 943 Ruban A. V (2016) Nonphotochemical chlorophyll fluorescence quenching: 944 mechanism and effectiveness in protecting plants from photodamage. Plant 945 physiology 170, 1903–1916. 946 Ruban A. V, Johnson M.P. & Duffy C.D.P. (2012) The photoprotective 947 molecular switch in the photosystem II antenna. Biochimica et Biophysica 948 Acta 1817, 167–181. 949 Smith P., Bustamante M., Ahammad H., Clark H., Dong H., Elsiddig E.A., … 950 Tubiello F.N. (2014) Agriculture, Forestry and Other Land Use (AFOLU). 39 951 Climate Change 2014: Mitigation of Climate Change. Contribution of 952 Working Group III to the Fifth Assessment Report of the Intergovernmental 953 Panel on Climate Change, 811–922. 954 Smith W.K., Biederman J.A., Scott R.L., Moore D.J.P., He M., Kimball J.S., … 955 Litvak M.E. (2018) Chlorophyll fluorescence better captures seasonal and 956 interannual gross primary productivity dynamics across dryland 957 ecosystems of southwestern North America. Geophysical Research Letters 958 45, 748–757. 959 Sonoike K. (2011) Photoinhibition of Photosystem I. Physiologia Plantarum 142, 960 56–64. 961 Soukupová J., Cséfalvay L., Urban O., Košvancová M., Marek M., Rascher U. & 962 Nedbal L. (2008) Annual variation of the steady-state chlorophyll 963 fluorescence emission of evergreen plants in temperate zone. Functional 964 Plant Biology 35, 63–76. 965 Springer K.R., Wang R. & Gamon J.A. (2017) Seasonal patterns of 966 photosynthesis, fluorescence, and reflectance indices in boreal trees. 967 Remote Sensing 9, 1–18. 968 Sun Y., Frankenberg C., Jung M., Joiner J., Guanter L., Köhler P. & Magney T. 969 (2018) Overview of Solar-Induced chlorophyll Fluorescence (SIF) from the 970 Orbiting Carbon Observatory-2: Retrieval, cross-mission comparison, and 971 global monitoring for GPP. Remote Sensing of Environment 209, 808–823. 972 Sun Y., Frankenberg C., Wood J.D., Schimel D.S., Jung M., Guanter L., … 973 Yuen K. (2017) OCO-2 advances photosynthesis observation from space 974 via solar-induced chlorophyll fluorescence. Science 358. 40 975 Thum T., Zaehle S., Köhler P., Aalto T., Aurela M., Guanter L., … Markkanen T. 976 (2017) Modelling sun-induced fluorescence and photosynthesis with a land 977 surface model at local and regional scales in northern Europe. 978 Biogeosciences 14, 1969–1987. 979 van der Tol C., Berry J.A., Campbell P.K.E. & Rascher U. (2014) Models of 980 fluorescence and photosynthesis for interpreting measurements of solar- 981 induced chlorophyll fluorescence. Journal of Geophysical Research: 982 Biogeosciences 119, 2312–2327. 983 Verhoeven A. (2014) Sustained energy dissipation in winter evergreens. New 984 Phytologist 201, 57–65. 985 Verrelst J., Rivera J.P., van der Tol C., Magnani F., Mohammed G. & Moreno J. 986 (2015) Global sensitivity analysis of the SCOPE model: What drives 987 simulated canopy-leaving sun-induced fluorescence? Remote Sensing of 988 Environment 166, 8–21. 989 Verrelst J., van der Tol C., Magnani F., Sabater N., Rivera J.P., Mohammed G. 990 & Moreno J. (2016) Evaluating the predictive power of sun-induced 991 chlorophyll fluorescence to estimate net photosynthesis of vegetation 992 canopies: a SCOPE modeling study. Remote Sensing of Environment 176, 993 139–151. 994 Walther S., Voigt M., Thum T., Gonsamo A., Zhang Y., Kohler P., … Guanter L. 995 (2016) Satellite chlorophyll fluorescence measurements reveal large-scale 996 decoupling of photosynthesis and greenness dynamics in boreal evergreen 997 forests. Global Change Biology 22, 2979–2996. 998 Wellburn A.R. (1994) Spectral determination of chlorophylls a and b, as well as 41 999 total carotenoids, using various solvents with spectrophotometers of 1000 different resolution. Journal of Plant Physiology 144, 307–313. 1001 Wieneke S., Burkart A., Cendrero-Mateo M.P., Julitta T., Rossini M., Schickling 1002 A., … Rascher U. (2018) Linking photosynthesis and sun-induced 1003 fluorescence at sub-daily to seasonal scales. Remote Sensing of 1004 Environment 219, 247–258. 1005 Van Wittenberghe S., Alonso L., Verrelst J., Moreno J. & Samson R. (2015) 1006 Bidirectional sun-induced chlorophyll fluorescence emission is influenced 1007 by leaf structure and light scattering properties - A bottom-up approach. 1008 Remote Sensing of Environment 158, 169–179. 1009 Wong C.Y.S. & Gamon J.A. (2015) The photochemical reflectance index 1010 provides an optical indicator of spring photosynthetic activation in 1011 evergreen conifers. New Phytologist 206, 196–208. 1012 Yang P. & van der Tol C. (2018) Linking canopy scattering of far-red sun- 1013 induced chlorophyll fluorescence with reflectance. Remote Sensing of 1014 Environment 209, 456–467. 1015 Yang P., van der Tol C., Verhoef W., Damm A., Schickling A., Kraska T., … 1016 Rascher U. (2018) Using reflectance to explain vegetation biochemical and 1017 structural effects on sun-induced chlorophyll fluorescence. Remote Sensing 1018 of Environment. 1019 Zarco-Tejada P.J., Catalina A., González M.R. & Martín P. (2013) Relationships 1020 between net photosynthesis and steady-state chlorophyll fluorescence 1021 retrieved from airborne hyperspectral imagery. Remote Sensing of 1022 Environment 136, 247–258. 42 1023 Zarco-Tejada P.J., González-Dugo M. V. & Fereres E. (2016) Seasonal stability 1024 of chlorophyll fluorescence quantified from airborne hyperspectral imagery 1025 as an indicator of net photosynthesis in the context of precision agriculture. 1026 Remote Sensing of Environment 179, 89–103. 1027 Zarter C.R., Adams W.W., Ebbert V., Cuthbertson D.J., Adamska I. & Demmig- 1028 Adams B. (2006a) Winter down-regulation of intrinsic photosynthetic 1029 capacity coupled with up-regulation of Elip-like proteins and persistent 1030 energy dissipation in a subalpine forest. New Phytologist 172, 272–282. 1031 Zarter C.R., Demmig-Adams B., Ebbert V., Adamska I. & Adams W.W. (2006b) 1032 Photosynthetic capacity and light harvesting efficiency during the winter-to- 1033 spring transition in subalpine conifers. New Phytologist 172, 283–292. 1034 Zhang Y., Guanter L., Joiner J., Song L. & Guan K. (2018) Spatially-explicit 1035 monitoring of crop photosynthetic capacity through the use of space-based 1036 chlorophyll fluorescence data. Remote Sensing of Environment 210, 362– 1037 374. 1038 Zuromski L.M., Bowling D.R., Köhler P., Frankenberg C., Goulden M.L., 1039 Blanken P.D. & Lin J.C. (2018) Solar-induced fluorescence detects 1040 interannual variation in gross primary production of coniferous forests in the 1041 western United States. Geophysical Research Letters 45, 7184–7193. 1042 43 1043 Figure legends: 1044 Fig. 1. Measurement rationale during the spring recovery of 2015 in Pinus sylvestris. 1045 Fig. 2. Seasonal variation of meteorological and photosynthetic parameters. 1046 Fig. 3. Seasonal variation of foliar pigment content. 1047 Fig. 4. Seasonal variation in spectral chlorophyll fluorescence (ChlF). 1048 Fig. 5. Principal component analysis (PCA) for spectral chlorophyll fluorescence. 1049 Fig. 6. Pearson correlation (r) matrix between key study variables. 1050 Fig.7. The linear regression between individual spectral fluorescence from 660 to 780 nm 1051 (Fλ) and photosynthetic parameters (α and β). 1052 1053 Supporting Information 1054 Figure legends: 1055 Fig. S1. Seasonal changes in CO2 flux, α, β, αn, and βn, for each individual chamber. 1056 Fig. S2. Seasonal variation of leaf total PAR absorption (AT(PAR)). 1057 Fig. S3. Linear regression correlations of field FV/FM (FMS-2) with field FV/FM (MONI-PAM) 1058 and lab FV/FM (FMS-2). 1059 Fig. S4. The significance matrix of Pearson correlation (P) between key study variables. 1060 Fig. S5. Coefficients of determination of the linear regression of individual spectral ChlF with 780 1061 ΦF and integral spectral ChlF (∫660 𝐹 𝑑𝜆). 44 1062 1063 Figure 1. Measurement rationale during the spring recovery of 2015 in Pinus sylvestris. In 1064 situ measurements encompassed (1) continuous gas exchange measurements with 1065 automated chambers (N=2-4 trees) to estimate CO2 flux and derive the seasonal 1066 photosynthetic parameters α and β, and (2) continuous PAM fluorescence measurements 1067 (N=4) to derive maximum quantum yield of PSII (FV/FM), quantum yield of fluorescence (ФF), 1068 and sustained photochemical (PQS) and non-photochemical quenching (NPQS). For the lab 1069 measurement, needles from top branches (N=5) were sampled during each measuring day. 1070 (1) One portion was immediately frozen in liquid nitrogen for posterior pigments analysis. (2) 1071 Another portion was rapidly brought to the laboratory, and used to measure total PAR 1072 absorption (AT(PAR)) and spectral fluorescence (Fλ) respectively (see also Materials and 1073 methods section for details). 45 1074 1075 Figure 2. Seasonal variation of meteorological and photosynthetic parameters. (a) Daily 1076 mean temperature (°C; black line) and mean midday photosynthetically active radiation 1077 between 11:00 and 14:00 (PAR; µmol m-2 s-1; grey line), (b) mean midday CO2 flux between 1078 11:00 and 14:00 (µg m-2 s-1), (c) proxies of maximum photosynthetic light-use efficiency (α, 1079 black line) and maximum photosynthetic rate (β, grey line) in prevailing conditions estimated 1080 from the gas exchange data, (d) daily maximum quantum yield of PSII (FV/FM, black line) 1081 and quantum yield of fluorescence (ΦF, grey line), and (e) sustained photochemical (PQS) 1082 and (f) non-photochemical (NPQS) quenching measured using MONI-PAM. Lines indicate 1083 the continuous measurement. Open points represent the dates when sampling took place. 1084 Solid black line indicates the first cold spell on March 22nd, and dotted black line indicates 1085 the second cold spell on April 16th. Points in b-f represent means of four biological replicates 1086 (N = 4). Shadows represent ±SE. 46 1087 1088 Figure 3. Seasonal variation of foliar pigment content. (a) Total chlorophyll (Chl; µg mgDW - 1089 1), (b) Chlorophyll a (Chl a; µg mgDW-1), (c) Chlorophyll b (Chl b; µg mgDW-1), (d) 1090 Chlorophyll a / Chlorophyll b (Chl a/b), (e) Total carotenoid (µg mgDW-1), and (f) 1091 Carotenoid/Chlorophyll ratio (Car/Chl). Solid black line indicates the first cold spell on March 1092 22nd, and the dotted black line indicates the second cold spell on April 16th. Points represent 1093 means of five biological replicates (N = 5). Error bars represent ±SE. 47 1094 1095 Figure 4. Seasonal variation of spectral chlorophyll fluorescence (ChlF). (a-c) The shape of 1096 spectral ChlF from 680 to 770nm (Fλ; mW m-2 sr-1 nm-1). The contour figure of (b) indicates 1097 the Fλ values. Two examples of spectral ChlF in (c) were measured on March 25th (early 1098 spring), and July 20th (summer). (d) Red ChlF at 690 nm (F690, filled circles) and far-red 780 1099 ChlF at 740 nm (F740, open circles), (e) integral ChlF from 660 to 780nm (∫660 𝐹 𝑑𝜆), and (f) 1100 fluorescence ratio (F690/F740). Solid and vertical black line indicates the first cold spell on 1101 March 22nd, and the dotted and vertical black line indicates the second cold spell on April 1102 16th. Shadows in c and error bars in d-f represent ±SE (N=5). 48 1103 1104 Figure 5. Principal component analysis (PCA) for spectral chlorophyll fluorescence. Panels 1105 indicate spectral variability (top panels) and seasonal patterns (bottom panels) of PC1 (a,b), 1106 PC2 (c,d) and PC3 (e,f) respectively. Solid and vertical black line indicates the first cold 1107 spell on March 22nd, and dotted and vertical black line indicates the second cold spell on 1108 April 16th. 49 1109 1110 Figure 6. Pearson correlation (r) matrix between key study variables. The r value is indicated 1111 as colour (positive correlations are shaded blue, and negative correlations are shaded 1112 orange) and number. 50 1113 1114 Figure 7. The linear regression between individual spectral fluorescence from 1115 660 to 780 nm (Fλ) and photosynthetic parameters (α, filled circles and β, open 1116 circles). (a) Coefficients of determination (R2), (b) slopes of α (β) vs. Fλ, (c) 1117 intercepts of Fλ vs. α (β). 1118 1119 51 1120 Supporting Information of 1121 Do all chlorophyll fluorescence emission wavelengths capture the spring 1122 recovery of photosynthesis in boreal evergreen foliage? 1123 1124 Running title: Spectral fluorescence in evergreens 1125 1126 Chao Zhang1, 2, 3, *, Jon Atherton1, Josep Peñuelas2, 3, Iolanda Filella2, 3, Pasi 1127 Kolari4, Juho Aalto4, 5, Hanna Ruhanen6, Jaana Bäck7, Albert Porcar-Castell1, * 1128 1129 1130 Figure legends: 1131 Fig. S1. Seasonal changes in CO2 flux, α, β, αn, and βn, for each individual 1132 chamber. 1133 Fig. S2. Seasonal variation of leaf total PAR absorption (AT(PAR)). 1134 Fig. S3. Linear regression correlations of field FV/FM (FMS-2) with field FV/FM 1135 (MONI-PAM) and lab FV/FM (FMS-2). 1136 Fig. S4. The significance matrix of Pearson correlation (P) between key study 1137 variables. 1138 Fig. S5. Coefficients of determination of the linear regression of individual spectral 780 1139 ChlF with ΦF and integral spectral ChlF (∫660 𝐹 𝑑𝜆). 52 1140 1141 Figure S1. Seasonal changes in (a) CO2 flux, (b) maximum photosynthetic light- 1142 use efficiency parameter α, (c) maximum photosynthetic rate parameter β, (d) 1143 normalized α or αn, and (e) normalized β or βn, for each individual chamber. Note 1144 the different temporal coverage for chambers. The grey shadow indicates the 1145 period used for the normalization. 53 1146 1147 Figure S2. Seasonal variation of leaf total PAR absorption (AT(PAR)). Solid and 1148 vertical black line indicates the first cold spell on March 22nd, and the dotted and 1149 vertical black line indicates the second cold spell on April 16th. Points represent 1150 means of five biological replicates (N = 5). Grey error bars represent ±SE. 54 1151 1152 Figure S3. Linear regression correlations of FV/FM measured in the field using 1153 FMS-2 with FV/FM measured using MONI-PAM in the field (filled circles) and using 1154 FMS-2 in the lab (open circles). Points represent means of four (filled circles) or 1155 five (open circles) biological replicates (N = 4 or 5). Dotted grey line is 1:1 line. 1156 ***P<0.001. Error bars represent ±SE (N = 4 or 5). 55 1157 1158 Figure S4. The significance matrix of Pearson correlation (P) between key study 1159 variables. The P value is indicated as colour and number. 56 1160 1161 Figure S5. Coefficients of determination (R2) of the linear regression of individual 1162 spectral fluorescence (Fλ) with (a) PAM-based fluorescence yield (ΦF) and (b) 780 1163 integral spectral ChlF from 660 to 780nm (∫660 𝐹 𝑑𝜆 57