Correction of PRI for carotenoid pigment pools improves photosynthesis estimation across different irradiance and temperature conditions

Highlights

• We studied PRI dynamics in changing irradiance and temperature setup.

• Sensitivity of PRI to LUE was dependent on the changing ratio of Chla+b/Carx+c.

• Interference from Chla+b/Carx+c in LUE estimation was lowered when using ΔPRI.

• Effect of low temperature on LUE estimation from PRI was reduced with ΔPRI.

• Leaf area index of trees appeared to be an important driver in observed relationships.

Abstract

We studied the influence of changing carotenoid pigments on the sensitivity of the photochemical reflectance index (PRI) to photosynthesis dynamics. The goal of the measurements was to examine how the introduction of ΔPRI into the working dataset can improve the estimation of photosynthesis. Spectral and photosynthetic characteristics of European beech and Norway spruce saplings were periodically measured in growth chambers with an adjustable irradiance and temperature. Patterns of environmental changes inside the growth chambers were created by periodic changes in irradiance and temperature. Four general irradiance periods lasting 10–12 days each were established. Introduced irradiance regimes varied in the sum of daily irradiance and amplitude of irradiance changes. Temperature was changed with more complex patterns to induce changes in xanthophyll cycle pigments at various time scales within these regimes. Our measurements confirmed the PRI linkage to photosynthetic light use efficiency (LUE). However, the strength of this connection was found to be dependent on changing pigment concentrations, specifically on the change in the ratio of chlorophylls to carotenoids. Furthermore, a negative interference in photosynthesis estimation from PRI was recorded if the temperature was lowered overnight to 12 °C. The differential PRI (ΔPRI), calculated as the simple difference between the PRI value measured during the daytime period and in early morning (PRI₀), revealed a decreased effect from pigments and cold temperature on LUE estimation. The regression analysis among all measured data identified an increased association between PRI and LUE following the introduction of ΔPRI from R² = 0.26 to 0.69 in beech and from R² = 0.61 to 0.77 in spruce data. The analyses showed that both leaf carotenoid concentrations and the conversion state of xanthophyll cycle pigments played a significant role in determining PRI and PRI₀ values and that the accurate assessment of these pigments in PRI across multiple levels of stress from irradiance and temperature might improve estimations of LUE through ΔPRI. In our data, ΔPRI appeared to be a good measure of photosynthesis, the dynamics of which differed between beech and spruce saplings upon switching temperatures.

Introduction

The usefulness of a photochemical reflectance index (PRI) approach to link plant photosynthetic CO₂ uptake efficiency in changing light (light use efficiency, LUE) and remotely sensed data has been widely reported (Garbulsky et al., 2008; Peñuelas et al., 2011). PRI is based on our understanding of the photoprotective role of xanthophyll cycle pigments within plants. When plants absorb more light energy than can be used by chlorophyll to produce glucose, excessive light energy is either transferred to xanthophyll molecules and emitted as heat or emitted as fluorescence (Gamon et al., 1992; Rascher et al., 2009). Variations in xanthophyll cycle pigment concentrations and conversions produce reflectance changes at a wavelength of 531 nm. Comparing this reflectance with a reference wavelength (typically 570 nm) can be used to detect stress (Gamon et al., 1997, Gamon et al., 1992; Peñuelas et al., 1995). Examination of empirical work has determined a fundamental relationship between plant photochemistry and PRI that is based on the responsiveness of photosynthesis to irradiance.

The variability of factors that may affect canopy-measured PRI can produce confusion in interpreting PRI, in addition to the desirable effects introduced by changing pigments. In field applications, sun–canopy–sensor geometry can exert a strong effect on the resulting PRI values, as light fields may vary in complex ways with canopy structure (Middleton et al., 2009; Sims et al., 2006). Wu et al. (2015) examined crops and found the structural dependence of PRI on leaf area index (LAI). In addition, these authors proposed a solution for removal of the structural signal, and their results imply that LAI change as a consequence of a change in the illumination angle can impact the measured PRI of the canopy. The results reported by Wu et al. (2015) and Gitelson et al. (2017) suggest that canopy structure-related LAI interferences should be considered when evaluating PRI responses between canopies with lower LAI.

The foliar ratio of chlorophylls to carotenoids (Chla+b/Carx+c) has a strong impact on the observed PRI variability (Filella et al., 2009). Decreasing Chla+b/Carx+c may be related to photosynthesis downregulation in stressed plants, but the changes in pigments may also cause impairment of the PRI–LUE relationship over the long term. Attempts have been made to eliminate the Chla+b/Carx+c variability in the measured signal by deconvoluting the influence of pigments. Gamon and Surfus (1999) have shown an opportunity to detract the extent of xanthophyll cycle pigment conversions by subtracting measured PRI value from the starting PRI value in introduced ΔPRI (ΔPRI = PRI₀ – PRI). Following this work, Gamon and Berry (2012) characterized differences in ΔPRI between sun and shade leaves as induced with pigment changes. This study has shown that PRI₀ represents a PRI state to which PRI values can be related if the comparison of diurnal changes and plant groups is required. Ideally, PRI₀ is measured at low irradiance on leaves with inactivated protective functions to isolate the slow changing component of PRI most likely related to Chla+b/Carx+c and the concentration of lutein pigments. Magney et al. (2016) determined that the ΔPRI does respond to diurnal physiological changes resulting from changes in VPD, air temperature, and stomatal conductance and suggested that these changes may be dependent on the observed pigment dynamics. Nitrogen availability affecting amounts of chlorophylls was considered a primary driver of ΔPRI sensitivity in this study, defining the slope of the observed dependency. ΔPRI yields better correlations in nutrient-deficient plots, thereby indicating the importance of carotenoid levels in observed relationships. There are several good examples showing the opportunity to better estimate LUE from PRI measured at the top of the canopy, if the value of PRI is corrected to the morning PRI₀ (Hmimina et al., 2015, Hmimina et al., 2014; Ripullone et al., 2011; Soudani et al., 2014). The presented studies often involve drought stress in the experimental design as the main factor driving photosynthesis declines and the magnitude of PRI response (Magney et al., 2016). We intend to further study the basics of the improved relationship between photosynthetic LUE and ΔPRI, and the desirable role of changing pigments in developing these relationships under conditions involving induced changes in irradiance and temperature.

Differences in the sensitivity of photosynthetic processes to light and temperature stress among species suggest the occurrence of interactive effects on photosynthetic pigments. An insufficient understanding of the associated reflectance signature in relation to the measure of photosynthesis is consequential (Ollinger, 2011; Sims and Gamon, 2002). The co-variation of environmental drivers in the upper canopy suggests that leaves in the upper canopy are often exposed to greater stress originating from exposure of leaves to direct light (Niinemets and Valladares, 2004). Upper canopy leaves may suffer from a variety of additional stresses, among which temperature stress may play a particularly substantial role (Williams et al., 1996). The upper canopy crown, usually accounted for in spectrometric measurements, may represent a substantial portion of the canopy involved in the observed gas exchange (Coops et al., 2017). However, the functional trait attributes of plants may not be the only features able to explain changes in PRI in a given environment. It may also be necessary to account for the PRI variations with changing Chla+b/Carx+c that determines the constitutive (slow-changing) component of the PRI. It has been suggested that the finer definition of the facultative process (fast-changing with xanthophyll cycle) can help to detect LUE and responses to summer stress, such as heat and drought (Gamon and Bond, 2013). Our previous development provided elementary knowledge of PRI changes in relation to dynamic fluctuations of photosynthetically active radiation (PAR) and temperature (Kováč et al., 2018). This study suggests large improvement in estimating LUE from ΔPRI, with measurements of PRI₀ that correspond to the xanthophyll cycle pigment conversion state in the dark. Changes in PRI–LUE and ΔPRI–LUE relationships have yet to be investigated with dynamic irradiance increases and temperature changes occurring on a daily scale.

We aimed to study further the sensitivity of both PRI and ΔPRI to decreasing and increasing photosynthesis in fluctuating environments. In this study, we focused on observations of the effect of mutual interactions among the plant pigments and overall canopy LAI on the developing relationship between PRI and LUE in changing temperature conditions. Understanding the aspects of LUE estimations using PRI would improve the applicability of PRI measurements in remote-sensing applications. We aimed to overcome the canopy structural effect on measured PRI by fixing the illumination–observation setup of the measurement by measuring canopies within the closed environment of a growth chamber. We thus compared responses in photosynthesis, pigments and PRI of tree species with distinctive sensitivities to low and high temperatures. Norway spruce trees, which prefer cold regions and higher altitudes, were compared with European beech trees, which show tolerance to higher temperatures and grow at lower altitudes. We established four regimes that varied in their daily sum of irradiance income even as the trees were undergoing similar periodic changes in temperature on a daily scale. We measured how the changes in pigments within these regimes affected the observed PRI–LUE relationship and examined the role of established pigment concentrations in the developing sensitivity of PRI and ΔPRI to LUE.