Time-integrated δ2H in n-alkanes and carbohydrates from boreal needles reveal intra-annual physiological and environmental signals

Abstract

Introduction

The stable H isotope ratio (δ²H) in plant-derived biomarkers (e.g. n-alkanes, tree-ring cellulose) has proven value, and profound potential, for insights into paleoenvironmental reconstructions and plant stress response (Dawson et al., 2004; Kahmen et al., 2013; Lehmann et al., 2024a). A key barrier that limits their reliability and implementation is that their H isotope composition can be affected by multiple isotope fractionating (Zhu et al., 2020; Holloway-Phillips et al., 2022; Baan et al., 2023a) and nonfractionating (i.e., mixing; Liu et al., 2021) processes. This can interfere with meaningful environmental or physiological insights from δ²H values in plant bioindicators (Baan et al., 2023b). Undoubtedly, they will benefit from a thorough examination of how the δ²H of plant compounds correlates with environmental and physiological signals under natural conditions.

Background

n-Alkanes with chain lengths of 25–35 carbons are straight-chained hydrocarbons found in leaf epicuticular waxes, which can be key contributors to the δ²H of n-alkanes (δ²H_alkane, Table 1) in soils, used to interpret past climate (Dawson et al., 2004; Schefuß et al., 2005; Thomas et al., 2021). Overall, δ²H_alkane acts as an indicator of δ²H in precipitation (δ²H_precip), modified by leaf evaporative enrichment, local meteorological factors (i.e. evapotranspiration (ET) and relative humidity (RH)) and source water (Sachse et al., 2006), which is further modified by biochemical isotope fractionation (Newberry et al., 2015; Baan et al., 2023a,b). δ²H_alkane in leaves can be related to δ²H in leaf water (δ²H_l-water; Freimuth et al., 2017; Zhu et al., 2020; Lehmann et al., 2024b), though not necessarily (McInerney et al., 2011), and evidence from seasonal δ²H_alkane variability of new needles in a natural forest shows that the δ²H_l-water signal can be obscured, potentially by changes in carbohydrate sourcing throughout the season (Newberry et al., 2015). Furthermore, the δ²H_alkane in new leaf tissue can record a shift from heterotrophy to autotrophy (Tipple & Ehleringer, 2018), and if leaves are sampled before autotrophy has been fully established, then the δ²H_l-water signal may be obscured (Zhu et al., 2020).

Table 1. Abbreviations and symbols used in the text.

Abbreviation	Description
NADPH	Nicotinamide adenine dinucleotide phosphate
NSC	Nonstructural carbohydrates
WSC	Water-soluble carbohydrates
δ2H	Isotope ratio of 2H compared to 1H, relative to VSMOW (‰)
δ2Halkane	δ2H of n-alkanes
δ2Hprecip	δ2H of precipitation
δ2Hl-water	δ2H of leaf water
δ2Hn-water	δ2H of needle water
δ2Hstarch	δ2H of starch
δ2HWSC	δ2H of water-soluble carbohydrates
δ2Hsource	δ2H of source water
δ2Hvapor	δ2H of water vapor
δ18O	Isotope ratio of 18O compared to 16O, relative to VSMOW (‰)
Δ2Hn-water	Needle water 2H enrichment above source (twig) water (‰)
ɛbio	Hydrogen isotope offset between n-alkanes and leaf water
RH	Atmospheric relative humidity
E	Transpiration rate
An	Net assimilation rate
gs	Stomatal conductance
T:RH	Time-integrated atmospheric relative humidity
T:E	Time-integrated transpiration rate
T:An	Time-integrated net assimilation rate
T:gs	Time-integrated stomatal conductance
T:δ2Hn-water	Time-integrated δ2H of needle water
T:∆2Hn-water	Time-integrated needle water 2H enrichment above source (twig) water
T:δ2Hsource	Time-integrated δ2H of source water
0N	Current-year needles
1N	One-year-old needles
R2(M)	Marginal R2. A pseudo-R2 estimate for the models being tested
R2(C)	Conditional R2. A pseudo-R2 estimate for the models being tested combined with model random effects, such as sampling date, time and site.
ICC	Intraclass correlation. The probability that two values from the same sampling date, and/or tree identity, correlate, on a scale of 0–1.

Interspecies differences in leaf δ²H_alkane are more constant between 2 yr than for leaf cellulose δ²H, suggesting that species differences in biosynthetic fractionation for δ²H_alkane are likely the more invariable and less sensitive to environmental variability (Baan et al., 2023a). Meanwhile, at the scale of intraleaf variability, δ²H_alkane is likely more environmentally sensitive than leaf cellulose δ²H, owing to a higher proportion of autotrophic production during n-alkane syntheses (Zhu et al., 2020). Leaf nonstructural carbohydrates, such as sugars (e.g. sucrose and glucose) and starch, are key intermediaries in the transfer of the δ²H signal from leaf water to cellulose (Holloway-Phillips et al., 2022; Lehmann et al., 2022), as observed for carbon-13 (δ¹³C) and oxygen-18 (δ¹⁸O) (Gessler et al., 2009). It is valuable to understand their role as δ²H intermediaries between leaf water and tree rings because tree-ring δ²H can correlate with δ²H_l-water (Lehmann et al., 2024a). Yet, prevailing intra-annual physiological or climatic signals from δ²H_WSC and δ²H_starch has not yet been explored in a natural forest. Evidence from a ²H-enrichment study shows that the δ²H in leaf sucrose is likely determined by physiological processes (Augusti et al., 2006). Further evidence shows the ²H offset between leaf sucrose and water having a variable relationship with δ²H_l-water, in addition to weak relationships with dark respiration, sugar pool turnover time and the proportion of sugar in the sugar and starch pool (Holloway-Phillips et al., 2022). These trends were reinforced by results from the δ²H of water-soluble carbohydrates (WSCs), a mixture of sugars and sugar alcohols, suggesting that δ²H in leaf sugars is determined by the relative concentrations of sugars and starch, and leaf gas exchange (Lehmann et al., 2024b). It is thus valuable to observe whether these signals exhibited in glasshouse experiments (Holloway-Phillips et al., 2022; Lehmann et al., 2024b) can also be observed in intra-annual variation of a natural forest. Water-soluble carbohydrates are used as a proxy for leaf sugars (Leppä et al., 2022; Tang et al., 2022; Lehmann et al., 2024b), as the simplest bulk matter extract that can be measured without using compound-specific isotope analysis. We hereafter use leaf WSC δ²H (δ²H_WSC) as a proxy for leaf sugar δ²H. This comes at the disadvantage of measuring a diverse group of leaf compounds with different metabolic history.

Even though leaf carbon partitioning between sucrose and starch is likely instrumental to δ²H in sugars and tree-ring cellulose (Holloway-Phillips et al., 2022; Wieloch et al., 2022; Lehmann et al., 2024a), the δ²H of starch (δ²H_starch) has not been compared with physiological or environmental trends. δ²H_starch is distinctly different from δ²H of sugars (Schleucher et al., 1999); therefore, it is important to investigate whether δ²H_starch has different physiological or climatic trends compared with δ²H in sugars because the temporally variable interaction between the starch and sugar pool may introduce mixed physiological or environmental signals absent from freshly assimilated sugars.

Hydrogen atoms in leaf carbohydrates (e.g. sugars and starch) and n-alkanes are derived from the same leaf's water, but their δ²H are different, owing to their distinct biochemical pathways (Schleucher et al., 1999). For example, among the prominent hydrogen isotope-fractionating processes that can lead to their different δ²H (Lehmann et al., 2024b), they can source different amounts of H atoms from nicotinamide adenine dinucleotide phosphate (NADPH), and these H atoms from NADPH can be a product of different NADPH-synthesis pathways (Sessions et al., 1999; Zhou et al., 2016; Wijker et al., 2019). Furthermore, their precursor molecule sources can vary between photosynthetic and glycolytic origins and differently represent short-term and long-term storage (Cormier et al., 2018; Zhu et al., 2020; Lehmann et al., 2024b). Further, nonfractionating (i.e., mixing) processes likely further affect carbohydrate and n-alkane δ²H differently, owing to their different metabolic fates – and these are particularly relevant in field studies (e.g. Leppä et al., 2022). The combined consequences of isotope-fractionating and nonfractionating processes, to δ²H_alkane and leaf carbohydrate δ²H, are poorly understood. Therefore, it is highly valuable to investigate δ²H_alkane and leaf carbohydrate δ²H together, to elucidate their physiological and environmental information, enhancing our understanding of the relative importance of their different drivers.

Temporal integration perspectives

A key step towards elucidating intra-annual physiological and environmental signals from leaf δ²H_alkane, δ²H_WSC and δ²H_starch in situ is quantifying important temporal aspects that could otherwise interfere with finding prevailing trends. Time integration has been long-established in the interpretation of δ¹⁸O in organic compounds, especially tree rings (Gessler et al., 2009; Pérez-de-Lis et al., 2022), and it is known to be important for the transfer of the δ¹⁸O signal from leaf water to foliar water-soluble organic matter (Barnard et al., 2007), WSC (Leppä et al., 2022) and resin (Tang et al., 2024). Therefore, it is long overdue to determine whether time-integrated, intraseasonal signals are revealed from δ²H_alkane, δ²H_WSC and δ²H_starch. This is outstandingly relevant for δ²H_alkane because n-alkanes are gradually accumulated during leaf development (Jetter et al., 2000). Indeed, leaf wax measurements from the evergreen shrub Prunus laurocerasus have shown that the leaf wax layer continues to thicken after leaf expansion; there were 10 molecular layers in 10-d-old leaves during expansion, c. 20 molecular layers after 50 d, and 1-yr-old leaves had 35–45 layers (Jetter et al., 2000; Jetter & Schäffer, 2001). Given that new leaf tissue growth coincides with the predominant wax production, it is a reasonable assumption that the leaf δ²H_alkane largely represents only the early stages of the leaf lifespan (Kahmen et al., 2011; Sachse et al., 2015; Freimuth et al., 2017). Other recent studies argue for the continuous formation of n-alkanes in leaves during the growing season (Speckert et al., 2023) and rapid response of n-alkane formation as a response to environmental stress such as drought, without increased concentration of n-alkanes in the wax layer (Srivastava & Wiesenberg, 2018). Since temperature and elevated CO₂ can affect the composition of n-alkanes and their precursors, fatty acids, within leaves (Ofiti et al., 2023), the hydrological signal in δ²H_alkane may not only be obscured by integration time but also the variability in conditions exposed to leaves during their integration.

Evidence from WSC δ¹⁸O would suggest that the δ²H_WSC integration period during a growing season can vary from within 48 h to more than 5 d (Leppä et al., 2022). Meanwhile, the δ²H_starch integration period could depend on the relative contribution of transitory starch to the total leaf starch pool. If the leaf starch pool is mostly transitory, the δ²H_starch integration period will likely be 1 d (Weise et al., 2011; Fernandez et al., 2017).

Spatiotemporal integration perspectives

δ²H_l-water becomes higher with increased proximity to evaporative sites (Luo & Sternberg, 1992; Gan et al., 2002; Farquhar & Gan, 2003). This phenomenon has demonstrated relevance for leaf spatial patterns in δ²H of cellulose and δ²H_alkane (Zhu et al., 2020; Liu et al., 2021). Since photosynthetic tissue water can be more exposed to evaporative ²H enrichment than bulk δ²H_l-water, leaf water isotope heterogeneity could be relevant for leaf δ²H_WSC, δ²H_starch and δ²H_alkane, like it has been reported for δ¹⁸O in sucrose (Baca Cabrera et al., 2023). Its role could be profound because the δ²H of source water (δ²H_source) and evaporative ²H enrichment that make up δ²H_l-water are distinctly different (Cernusak et al., 2016); therefore, an increased ratio of evaporative ²H enrichment could substantially change the δ²H of surrounding water during n-alkane and carbohydrate syntheses. Resultantly, it is not clear whether the long-term δ²H_WSC, δ²H_starch and δ²H_alkane signals are more reflective of needle water δ²H (δ²H_n-water) or the ²H enrichment of needle water above source water (∆²H_n-water). Investigating the relative contributions of δ²H_n-water and ∆²H_n-water to δ²H_WSC, δ²H_starch and δ²H_alkane in natural environments is crucial because it determines the ecohydrological conditions that they represent during palaeohydrological reconstructions.

Study aim and hypotheses

We investigated whether there are time-integrated, physiological or environmental signals in the variability of δ²H_alkane, δ²H_WSC and δ²H_starch using a comprehensive seasonal field survey of Pinus sylvestris L. (Scots Pine). We hypothesized that (1) δ²H_alkane and δ²H_WSC have stronger relationships to physiological (leaf gas exchange) or environmental factors (e.g. RH and modeled water isotopes) after accounting for multiple integration days or weeks, and (2) δ²H_alkane, δ²H_WSC and δ²H_starch are more strongly related to time-integrated ∆²H_n-water than δ²H_n-water because they can be sensitive to inhomogeneities in leaf water (Zhu et al., 2020; Liu et al., 2021), and photosynthetic tissue water is likely more exposed to evaporative enrichment than bulk leaf water (Baca Cabrera et al., 2023). Finally, we hypothesized that (3) based on evidence that δ²H_alkane is related to δ²H_l-water in controlled conditions (Freimuth et al., 2017; Lehmann et al., 2024b), enhanced by high autotrophic production in the field (Zhu et al., 2020), δ²H_alkane will reflect δ²H_l-water while δ²H_WSC and δ²H_starch will represent leaf gas exchange (Holloway-Phillips et al., 2022; Lehmann et al., 2024b).