Research Collection

Doctoral Thesis

Partial transfer of leaf water ²H-enrichment and variable biosynthetic fractionation affect the leaf wax n-alkane δ²H values in grasses

Author(s): Gamarra, Bruno

Publication Date: 2016

Permanent Link: https://doi.org/10.3929/ethz-a-010713453

Rights / License: In Copyright - Non-Commercial Use Permitted

This page was generated automatically upon download from the ETH Zurich Research Collection. For more information please consult the Terms of use.

ETH Library DISS. ETH No. 23286

Partial transfer of leaf water 2H-enrichment and variable biosynthetic fractionation

affect the leaf wax n-alkane δ 2H values in grasses

A thesis submitted to attain the degree of

DOCTOR OF SCIENCES of ETH ZURICH

(Dr. sc. ETH Zurich)

presented by

BRUNO GAMARRA

MSc ETH Environ. Sc., ETH Zurich

born on 23.06.1982

of Peruvian nationality

accepted on the recommendation of

Prof. Dr. Ansgar Kahmen

Prof. Dr. Tim Eglinton

Dr. Rolf Siegwolf

2016

Table of contents i

Table of contents

Summary ...... 1

Zusammenfassung ...... 5

Chapter 1 ...... 9

General introduction

Chapter 2 ...... 25

Effects of leaf water evaporative 2H-enrichment and biosynthetic fractionation on leaf wax n-alkane δ2H values in C3 and C4 grasses

Chapter 3 ...... 67

Low secondary leaf wax n-alkane synthesis on fully mature leaves of C3 grasses grown at low and high humidity

Chapter 4 ...... 95

Concentrations and δ2H values of cuticular n-alkanes vary significantly among plant organs, species and habitats in grasses from an alpine and a temperate European grasslands

Chapter 5 ...... 129

Conclusions

Acknowledgements ...... 137 ii Table of contents

Summary 1

Summary

Leaf wax n-alkanes are long chained hydrocarbons contained in the cuticle of terrestrial plants. Leaf wax n-alkane δ2H values have been successfully used to study the hydrological cycle, plant-water and plant-carbon relationships. However, the physiological and biochemical drivers that shape leaf wax n-alkane δ2H values are not completely understood.

It is particularly unclear, why n-alkanes in grasses are typically 2H-depleted compared to dicotyledonous plants and why C3 grasses are 2H-depleted compared to C4 grasses. Also, the timing of n-alkane synthesis and the de novo synthesis of n-alkanes in fully matured leaves are still matter of debate. On top, calibration studies designed to resolve sources of heterogeneity in n-alkane δ2H values have exclusively focused on n-alkanes derived from leaves.

To solve these uncertainties three studies aimed to explore how leaf wax n-alkane δ2H values in grasses are influenced by plant physiological and biochemical processes. In the first study the effect of leaf water evaporative 2H-enrichment (LW Δ2H) on n-alkane δ2H values was quantified for a range of C3 and C4 grasses that were grown in climate chambers under controlled environmental conditions. It was found that only a fraction of the LW Δ2H is imprinted on the n-alkane δ2H values in C3 and C4 grasses. It was also detected that the biosynthetic hydrogen isotope fractionation (εbio) was different for C3 and C4 grasses. As such, differences between leaf wax n-alkane δ2H values of C3 and C4 grasses would not be

2 driven by LW Δ H but largely the result of systematic differences in εbio between these two plant groups. 2 Summary

In the second study the timing of the leaf wax n-alkane synthesis was investigated.

Specifically the secondary leaf wax n-alkane synthesis was tested in mature leaf blades of C3 grass species. The experiments showed an incorporation of hydrogen from highly 2H- enriched irrigation water into the wax n-alkanes from mature leaves. Mature grass leaves continued the synthesis of wax n-alkanes after leaf emergence. The rate of secondary n- alkanes synthesis was, however, relatively low and varied among species from 0.09% to

1.09% per day. As such, leaf wax n-alkane δ2H values would be determined mainly by environmental and physiological conditions in the beginning of the life of a leaf.

In the third study n-alkane concentration and δ2H values of different grass organs were surveyed. Leaves, sheaths, stems, inflorescences and roots were sampled from a total of 15 species of C3 grasses in temperate and alpine grasslands in Switzerland. It was detected that inflorescences had typically much larger n-alkane concentrations compared to other organs while roots had very low n-alkane concentrations. The δ2H values of the carbon autonomous plant organs such as leaves, sheaths and stems were in general more negative compared to the non-carbon autonomous organs such as inflorescences and roots. Variable n-alkane δ2H values in different plant organs could be the result of differences in the H-

NADPH biosynthetic origin in response to the carbon autonomy of the plant organ.

Overall, this thesis brings new insights into the natural variability of leaf wax n-alkane

δ2H values in grasses. The incomplete transfer of LW Δ2H to the leaf wax n-alkanes δ2H values in grasses can explain why grasses are typically 2H-depleted compared to dicotyledonous plants. The low secondary leaf wax n-alkane synthesis in grass leaves after maturity suggests that in general leaf wax lipid δ2H values do not record environmental and plant physiological processes beyond leaf maturity. Finally, a different εbio between C3 and

C4 grasses and also between grass organs suggests that n-alkane δ2H values have a great Summary 3

potential as indicator of changes in plant carbon autonomy. This has important implications for the interpretation of n-alkane δ2H values in plant physiology and paleoecology.

4 Summary

Zusammenfassung 5

Zusammenfassung

Blattwachs n-Alkane sind langkettige Kohlenwasserstoffe die in der Epidermis von

Landpflanzen vorkommen. δ2H-Werte von Blattwachs n-Alkanen wurden bisher erfolgreich verwendet um den Wasserkreislauf und die Beziehungen zwischen der Pflanze und dem

Wasser sowie dem Kohlenstoff zu untersuchen. Dennoch sind die Auswirkungen von physiologischen und biochemischen Prozessen auf die δ2H-Werte von Blattwachs n-Alkanen bisher wenig verstanden. Vor allen Dingen ist unklar, warum die n-Alkane in Gräsern, verglichen zu dikotylen Pflanzen, grundsätzlich in 2H abgereichert sind, und warum C3 Gräser stärker abgereichert sind als C4 Gräser. Auch ist der Zeitpunkt der n-Alkan Synthese und der de novo Synthese von n-Alkanen in voll entwickelten Blättern immer noch ein

Diskussionsgegenstand. Bisher veröffentlichte Kalibrationsstudien, welche die Ursachen der

Heterogenität von n-Alkan δ2H-Werten untersuchten, fokussierten dabei ausschliesslich auf die n-Alkane von Blättern.

Für eine genauere Erforschung dieses Themas, wurden in dieser Doktorarbeit drei

Studien durchgeführt, welche die Effekte von physiologischen und biochemischen Prozessen auf die δ2H-Werte von Blattwachs n-Alkanen in Gräsern untersuchen. In der ersten Studie wurde der Effekt der evaporativen Blattwasser-2H-Anreicherung (LW Δ2H) auf die δ2H-Werte der n-Alkane für eine Auswahl von C3 und C4 Gräsern bestimmt. Dafür wurden die Gräser unter kontrollierten Umweltbedingungen in Klimakammern kultiviert. Die Studie ergab, dass die LW Δ2H-Anreicherung nur teilweise in den δ2H-Werten der n-Alkane der C3 und C4

Gräser sichtbar ist. Darüber hinaus wurde auch gezeigt, dass sich die biosynthetische

Fraktionierung der Wasserstoffisotopen (εbio) in C3 und C4 Pflanzen unterscheidet. Aus 6 Zusammenfassung

diesem Grund wird angenommen, dass die δ2H-Werte der Blattwachs n-Alkane von C3 und

C4 Gräsern nicht durch die LW Δ2H-Anreicherung bestimmt werden, sondern vielmehr durch systematische Unterschiede dieser beiden Pflanzengruppen in ihrer εbio.

In der zweiten Studie lag der Fokus auf der Bestimmung des Zeitpunktes der

Blattwachs n-Alkan Synthese in Gräsern. Dafür wurde die sekundäre Blattwachs n-Alkan

Synthese in voll entwickelten Blattspreiten von C3 Gräsern untersucht. Dabei konnten die

Experimente zeigen, dass Wasserstoff aus stark 2H-angereichertem Bewässerungswasser in

Wachs n-Alkane von voll entwickelten Blättern eingebunden wurde. Dabei fand die Wachs n-

Alkan Synthese in voll entwickelten Blättern auch nach dem Blattaustrieb weiterhin statt.

Die Rate der sekundären n-Alkan Synthese war jedoch relativ tief und variierte für verschiedene Arten zwischen 0.09 und 1.09% pro Tag. Aufgrund dieser tiefen Rate werden die δ2H-Werte der Blattwachs n-Alkane vor allem durch die Umweltbedingungen in der

Austriebsphase der Blätter festgelegt.

In der dritten Studie wurden die n-Alkan Konzentrationen und die δ2H-Werte von verschiedenen Grasorganen untersucht. Dafür wurden Blätter, Blattscheiden, Stängel,

Blütenstände und Wurzeln von 15 verschiedenen C3 Grasarten von temperierten und alpinen Grasflächen in der Schweiz gesammelt. Es wurde beobachtet, dass die Blütenstände generell viel höhere n-Alkan Konzentrationen aufwiesen als die anderen untersuchten

Organe. Im Gegensatz dazu hatten Wurzeln sehr niedrige n-Alkan Konzentrationen. Die δ2H-

Werte der Kohlenstoff-autonomen Pflanzenorgane wie Blätter, Blattscheiden und Stängel sind im Allgemeinen stärker negativ als die nicht Kohlenstoff-autonomen Pflanzenorgane wie Blütenstände und Wurzeln. Die Variabilität der n-Alkan δ2H-Werte in den verschiedenen

Pflanzenorganen könnte durch die unterschiedliche Herkunft des biosynthetischen H-NADPH verursacht sein, welches wiederum in Abhängigkeit zu der Kohlenstoff-Autonomität des

Pflanzenorgans steht. Zusammenfassung 7

Diese Doktorarbeit liefert neue Erkenntnisse über die natürliche Variabilität von δ2H-

Werten in Blattwachs n-Alkanen in Gräsern. Der unvollständige Transfer von LW Δ2H auf die

δ2H-Werte von Blattwachs n-Alkanen in Gräsern kann erklären warum Gräser im Vergleich zu dikotylen Pflanzen typischerweise in 2H abgereichert sind. Die schwache sekundäre

Blattwachs n-Alkan Synthese in voll entwickelten Blättern deutet darauf hin, dass die δ2H-

Werte von Blattwachslipiden im Allgemeinen keine über die Blattreife hinaus integrierte

Information über Umwelt- und pflanzenphysiologische Prozesse liefern. Letztendlich weisen

die unterschiedlichen εbio-Werte von C3 und C4 Gräsern sowie von verschiedenen

Pflanzenorganen auf ein grosses Potenzial der n-Alkan δ2H-Werte als Indikator für die

Kohlenstoff-Autonomität von Pflanzen hin. Dies hat eine wichtige Bedeutung für die

Interpretation von n-Alkan δ2H-Werten in der Pflanzenphysiologie und Paläoökologie.

Zusammenfassung 8

Chapter 1: General introduction 9

Chapter 1

General introduction

Global climate change has been regarded as influencing a vast part of earth natural processes, from raising the global temperature to changing the hydrological cycle (IPCC,

2014). Understanding how a changing environment affects forests and grasslands is a key question, because changes in plant physiological processes such as photosynthesis, growth or transpiration determine how an ecosystem will acclimate to a warmer or drier environment. However, there is yet insufficient understanding of such plant ecophysiological processes, especially related with water-plant relations, usually because projects with short experimental timescales from weeks to months fail in capturing long-term natural processes and feedbacks (Leuzinger et al., 2011).

In the last decades the use of stable isotopes in plant material has become a powerful tool to study environmental and plant physiological changes with large spatial and temporal resolution. Isotopes are atoms of the same element that differ in the number of neutrons.

Studies in plant ecology have used either stable isotopes occurring at natural abundances or at enriched levels. Natural abundances are usually used in ecology as integrators of natural processes or as tracers that follow the fate and transformations of certain compounds. In ecological studies isotope fractionation is used to investigate such processes and compounds. Isotope fractionation is defined as the change in the ratio between the light and heavy isotopes between a source and a substrate. Isotope fractionations occur because physical or chemical reactions favor the use of one isotope over the other when it is energetically convenient. Stable isotope ratios are expressed in the delta (δ) notation and 10 Chapter 1: General introduction

shown in per mil (‰) units as ratio of the light over the heavy isotope relative to the ratio of an international standard according to the following equation where ‘E’ is the studied element and ‘X’ is the mass of the heavy isotope:

�!"#$%& � !� = 1000 × − 1 , ‰ �!"#$%#&%

1.1. Stable isotopes from inorganic and organic material

Hydrological studies have classically used the relative abundance of hydrogen and oxygen isotope ratios in precipitation to relate with fluxes in the hydrological cycle (Craig and

Gordon, 1965; Gat, 1996). Natural waters, e.g. from precipitation, are stored in geological archives such as ice cores. Ice cores thus record long-term changes in precipitated δ2H and

δ18O values and provide important information for the reconstruction of the past and present hydrological cycle (Thompson et al., 2003). The use of continental ice cores is however constrained by their global availability, as they are restricted to polar and high mountain regions. This diminishes our understanding of the past hydrological cycle and its linkages with global climate and terrestrial ecology. Apart from ice cores, stable isotopes from organic material are widely used as a tool in environmental studies (Dawson et al.,

2002). Specifically, hydrogen and oxygen isotopes in plant material are widely used as they record δ2H and δ18O values from precipitation. It has also been observed that 2H, 13C and 18O isotopes of plant-originated material correlate with environmental variables such as temperature and with components of the hydrological cycle (Dawson and Ehleringer, 1993).

The stable isotope composition of plant material has thus been used with success as a proxy for environmental mechanisms. Isotope composition in organic material from sedimentary cores is also an important source of information in hydrological studies. Organic matter in lacustrine and sea sediments is largely composed of debris from photosynthetic organisms such as algae and cyanobacteria. Such organisms have in surface waters a principal source of Chapter 1: General introduction 11

hydrogen in the synthesis of compounds during their lifespan. Sedimentary cores have thus information of water isotope ratios during photosynthesis. As such, analysis of the isotope composition of sedimentary records is important in the study of the past hydrological cycle

(Estep and Hoering, 1981; Sternberg, 1988). The δ2H signal in sedimentary records is, however, difficult to interpret. This is because sedimentary records are a complex mixture of organic compounds with a distinctive hydrogen isotope signature. This isotope imprint depends on the type of photosynthetic organism, the complexity of its biochemical pathways and the exchange of hydrogen in secondary reactions in the environment

(Schimmelmann and Sessions, 2006). Therefore, isotope ratios derived from bulk organic matter in sedimentary cores are not always convenient to reconstruct past hydrological cycles because their use is obscured by uncertainties in the different factors that affect the isotopic composition of compounds that form part of such sedimentary records.

Stable isotopes in terrestrial plants are considered a powerful tool, not only for hydrological studies but also for understanding the relationships of plants and their surrounding environment (Dawson et al., 2002; Barbour, 2007). For example, δ2H and δ18O values from plant material e.g. cellulose, record information of isotopic precipitation but also contain information of plant evapotranspiration (Roden et al., 2000). As a consequence, analysis of isotope ratios from cellulose enables to have a robust mechanistic understanding of environmental and plant-physiological conditions at the moment of cellulose synthesis

(Treydte et al., 2006; Helliker and Richter, 2008; Kahmen et al., 2011b). However the use of cellulose as biomarker is constrained by degradation. Cellulose, as most other plant compounds, degrades and decays upon plant death. This means that the hydrogen and oxygen isotopes from cellulose only integrate information on a relatively small level, e.g. plant-water information of a plant but not an ecosystem, and can be used only in a relatively short period, i.e. few years to decades after plant death. In summary, δ2H and δ18O values from cellulose provide important information for paleohydrology and paleoecology however 12 Chapter 1: General introduction

their use has important limitations regarding the spatial and temporal integration of such information.

1.2. Hydrogen isotopic composition from plant cuticular lipids – leaf wax n-alkanes

Due to the recent instrumental development in isotope ratio mass spectrometry it is now possible to measure hydrogen and oxygen isotope ratios in specific compounds (Burgoyne and Hayes, 1998; Hilkert et al., 1999). In this respect, it is common to analyze the hydrogen of lipids such as esters, fatty acids, ketones, alcohols and alkanes, as they are present in a variety of organisms such as photosynthetic bacteria, algae and terrestrial plants. δ2H values of lipids from sedimentary records have shown to highly correlate with the δ2H values of the source waters where the photosynthetic organisms synthesize such lipids (Sessions et al.,

1999; Sauer et al., 2001; Chikaraishi et al., 2004; Sachse et al., 2004). Such correlations are nowadays used to successfully investigate the δ2H values of natural waters and to do paleohydrological reconstructions (Schefuß et al., 2005; Sachs et al., 2009). Water-lipid isotope correlations have however offsets. This suggests that environmental or physiological processes also affect the hydrogen isotope fractionation during lipid synthesis. As a consequence, it is still difficult to assess the effect of physiological processes on the lipid biomarker isotope composition. Understanding such effects would improve our interpretation of sedimentary biomarkers not only on a paleohydrological but also on a paleoecological and ecohydrological level (Krull et al., 2006).

Among other lipids it is particularly interesting the use of n-alkanes and their δ2H values. Leaf wax n-alkanes are long chained n-alkyl molecules with skeletons of 25 to 35 carbon atoms and are vital components of the plant cuticle of terrestrial plants. The ecological function of cuticular lipids, such as leaf wax n-alkanes, is to prevent the plant from losing water and as a mechanical barrier against environmental abrasion (Jetter et al., 2006). Chapter 1: General introduction 13

In n-alkanes the hydrogen is covalently bound to the carbon and does not exchange with hydrogen from the surrounding water (Sessions et al., 2004). n-Alkanes can therefore persist in the environment and geological records for millions of years after plant decay and surpass sedimentary diagenesis (Radke et al., 2005; Schimmelmann and Sessions, 2006). n-Alkanes are thus not only highly abundant in the wax cuticular layer of leaves but also in soils, sediments and even in the atmosphere (Eglinton and Hamilton, 1967). Leaf wax n-alkanes

δ2H values are not only advantageous in paleohydrological but also in ecological studies because they reflect source water δ2H values while at the same time plant-water relations e.g. leaf evapotranspiration (Sachse et al., 2006; Sachse et al., 2010; Feakins and Sessions,

2010). Additionally, the analysis of n-alkane δ2H values integrates information with tremendous spatial and temporal resolution. Those qualities make leaf wax n-alkanes and their δ2H values a powerful biomarker for hydrological, paleoclimatic and paleoecological studies (Eglinton and Eglinton, 2008).

1.3. Processes that affect the hydrogen isotopic composition of leaf wax n-alkanes

Although the use of leaf wax n-alkanes δ2H values has a great potential in paleohydrology, important information is intertwined within this isotopic signal. Recent studies suggest that this information is mainly ecophysiological (Smith and Freeman, 2006; Kahmen et al., 2013a and b). In essence, there are critical plant physiological mechanisms that determine the entire isotope signature of the leaf wax n-alkanes, which are still not completely understood.

These physiological mechanisms are relatively well determined for dicots. In dicots, the contribution from both, source water δ2H values and plant physiological mechanisms such as leaf transpiration on the isotope composition of leaf wax n-alkanes is well characterized. On the contrary, in monocots, the effects of plant physiological conditions such as leaf evaporative transpiration on their leaf wax n-alkanes δ2H values are quantitatively unclear. 14 Chapter 1: General introduction

Moreover, the interpretation of leaf wax n-alkane δ2H values is obscured by uncertainties in the hydrogen fractionation during biosynthesis. For example, it is unclear if and to which degree environmental variables such as light or temperature have an effect on the biosynthetic fractionation. Also, seasonal and across species variability in leaf wax n-alkanes

δ2H values suggests that plant functional groups or species-specific biochemical processes might also significantly affect this isotopic signal. Based on this ground, conceptual models suggest that source water (Sauer et al., 2001; Sessions, 2006), leaf water evaporative 2H- enrichment and biosynthetic fractionation (Chikaraishi and Naraoka, 2007) affect the leaf wax n-alkanes δ2H values. Figure 1.1 shows schematically some of the fractionation processes that have an influence on the leaf wax n-alkanes δ2H values. The contribution of each factor is still not completely characterized for grasses.

Leaf water

Leaf water

evapora3ve Biosynthe3c increasing 2H-enrichment water pool (LW Δ2H)

2 Source water H-enrichment

Biosynthe3c frac3ona3on

(εbio)

Leaf wax n-alkane

Figure 1.1. Factors that affect the leaf wax n-alkane δ2H values in grasses (Modified after Kahmen et al., 2013a) Chapter 1: General introduction 15

1.3.1. Source water

Because the principal source of hydrogen in plants is water, the δ2H values n-alkanes record the δ2H values of source water e.g. precipitation, and also contain information of the environmental variables affecting plant water sources e.g. soil evaporation (Sessions et al.,

1999; Sauer et al., 2001; Huang et al., 2004).

1.3.2. Leaf water evaporative 2H-enrichment

Water in the form of precipitation and surface waters is the principal source of hydrogen in the leaf water. Terrestrial plants, dicots or monocots, uptake water from different depths of the soil profile. Then, water is transported to the plant organs, a large fraction to the leaves.

Water is transpired out of the leaf due to evaporative processes, usually induced by a high

Vapor Pressure Deficit (VPD) between the air and the plant. The leaf evaporation favors the transpiration of water molecules containing hydrogen over the molecules with deuterium.

This phenomenon is defined as leaf water evaporative 2H-enrichment, as the leaf becomes water 2H-enriched (Roden and Ehleringer, 1999; Kahmen et al., 2008). As a consequence,

δ2H values from leaf water are usually more positive than their corresponding source water

δ2H values (figure 1.1). Alongside, the literature presents evidence that n-alkane δ2H values are shaped by a series of plant physiological factors occurring at the leaf level. For example, n-alkane δ2H values from terrestrial plants have been found to usually be higher than n- alkane δ2H values from aquatic plants (Sachse et al., 2004), even when both plants had the same resource water. This has been interpreted as an effect of leaf water evaporative 2H- enrichment because aquatic plants do not experience transpiration.

While different studies clearly show an effect of leaf water 2H-enrichment on n-alkane

δ2H values of dicots, this effect is not straightforward for grasses. This is complex because opposite to dicots, which entirely reflect leaf evaporative 2H-enrichment in their leaf wax n- alkanes δ2H values, grasses synthesize new leaf tissue in the meristem at the base of the 16 Chapter 1: General introduction

leaf. A sheath protects this basal part of the leaf from losing water and hence leaf wax n- alkanes δ2H values from monocots are only partially affected by leaf water evaporative 2H- enrichment. The effect of leaf water evaporative 2H-enrichment on leaf wax n-alkane δ2H values has been directly studied in grasses (McInerney et al., 2011; Kahmen et al, 2013a;b).

Those studies have shown opposite results. First, insensitivity of leaf wax n-alkane δ2H values to evapotranspiration was observed in a modeled approach (McInerney et al., 2011).

These results were later questioned by evidence that the hydrogen isotope ratio of n- alkanes in fact reflect leaf water evaporative enrichment (Kahmen et al., 2013a). This study, however, quantified this effect for only one C3 and C4 species.

1.3.3. Biosynthetic fractionation

The depletion of deuterium in n-alkane δ2H values compared to the water used for biosynthesis (εbio) has been characterized to have a broad range (Sessions et al., 1999; Smith and Freeman et al., 2006; Sachse et al., 2010) for a number of species and across different natural systems. The influence of environmental variables such as temperature or light on

εbio is still a matter of debate because no direct measurements or study has addressed this question. Recent studies have nevertheless associated changes in εbio with the source of hydrogen (Schmidt et al., 2003; Zhang et al., 2009). In essence, lipids and their biochemical precursors have two sources of hydrogen, one that comes directly from photosynthesis and another that comes from NADPH. Both have distinctive isotopic composition, more negative and more positive, respectively. Thus, n-alkane δ2H values would be controlled by a changing

εbio and its dependency on the origins of metabolic hydrogen, e.g. from photosynthesis or

NADPH. As such, n-alkane vales could indicate plant carbon autonomy, i.e. the use of recent photosynthesizes or carbon reserves.

Chapter 1: General introduction 17

1.3.4. Timing of n-alkane synthesis

A further effect that could affect leaf wax n-alkane δ2H values of grasses is the timing of lipid synthesis after leaf emergence and growth. The cuticle in leaves, which contains lipid waxes such as n-alkanes, is usually developed during its emergence, early in the life of the leaf

(Jetter and Schäffer, 2001). As such, δ2H values are established early during plant ontogeny and only environmental and plant physiological processes occurring during leaf emergence and development are recorded in the n-alkanes δ2H values. More recent evidence however suggests that cuticular lipids are synthesized constantly after leaf maturity, especially in the presence of environmental stressors (Jetter et al., 2006; Shepherd and Griffiths, 2006). Leaf wax n-alkane δ2H values could therefore record not only environmental and physiological variables until leaf maturity but also processes occurring after maturity during longer periods of plant growth. Seasonal heterogeneity in n-alkane δ2H values has been explained by different plant physiological conditions affecting the plant along a growth season (Smith and Freeman, 2006; Pedentchouk et al., 2008; Sachse et al., 2009; Sachse et al., 2010;

Feakins and Sessions, 2010). Additionally, species-specific differences in the timing of cuticle synthesis and stress-induced secondary cuticular synthesis obscure the interpretation of the leaf wax n-alkanes δ2H values. To date secondary synthesis of n-alkanes after the leaf maturity has only been studied in a few isotope-enriched experiments, with contradicting results. It has been indicated that grass leaves produce lipids, such as n-alkanes, regularly and complete lipid regeneration could take days to weeks (Gao et al., 2011). Opposing to that, other studies have found that n-alkanes are produced exclusively early during leaf growth and emergence in dicots (Kahmen et al., 2011a; Gao et al., 2012). It was however admitted that a secondary n-alkane synthesis could be triggered after environmental stressors, such as wind or environmental abrasion, reduce wax lipid abundances in the cuticle. 18 Chapter 1: General introduction

1.4. Scope and objectives of this thesis

This doctoral thesis focuses on investigating the effects of environmental and physiological processes on the δ2H values in leaf wax n-alkanes of C3 and C4 grasses. The revealing mechanistic understanding can be used to investigate past and present plant ecophysiological events using δ2H values of n-alkanes in living plants or plant archives.

Specifically, the following points are addressed:

• Effect of leaf water evaporative 2H-enrichment on leaf wax n-alkane δ2H values in C3

and C4 grasses

• εbio in C3 and C4 grasses

• n-Alkane synthesis after maturity of leaves in C3 grasses

• n-Alkane presence in other organs in C3 grasses

1.5. Outline

The objectives were investigated in three studies, which are further explained in the chapters 2, 3 and 4 of this thesis.

In chapter 2 the effect of leaf water evaporative 2H-enrichment on leaf wax n-alkanes

δ2H values is investigated. This effect is believed to be a key plant physiological mechanism in controlling the isotopic signatures of plant-derived n-alkanes in grasses. To quantitatively determine such an effect, five species of C3 and C4 grasses were grown under different humidity in Climate-controlled growth Chambers (CC). Also, transpiration rates were analyzed as an additional plant physiological process able to drive different magnitudes of such effect across species.

In chapter 3 the secondary leaf wax n-alkane synthesis after leaf maturity is studied. A secondary synthesis would imply that leaf wax n-alkane δ2H values record not only environmental and physiological conditions during short periods of leaf growth and Chapter 1: General introduction 19

emergence but also integrate such conditions in longer periods after leaf maturity. To trace secondary n-alkane synthesis, six species of C3 grasses were grown in CC. The isotopic composition of the irrigation water was changed to more 2H-enriched water after leaf maturity. δ2H values of mature leaves before and after change in irrigation water were analyzed.

In chapter 4 the δ2H values of different plant organs are investigated. Although leaf wax n-alkanes isotopic ratios are widely studied in leaves, there is not sufficient information of other plant organs, such as stems, inflorescences or roots. Different organ n-alkane δ2H values would suggest a different εbio and consequently hydrogen source acquisition. To analyze δ2H values of different plant organs, C3 grasses from two grasslands in the regions of

Ennetbaden (Aargau, Central Switzerland) and Preda (Graubünden, Eastern Switzerland) were sampled.

Finally, in chapter 5 the main findings of this thesis are presented. They are discussed in the context of plant physiological and paleoecological research. A final outlook is given for future studies in the area.

20 Chapter 1: General introduction

References

Barbour MM. 2007. Stable oxygen isotope composition of plant tissue: a review. Functional Plant Biology 34: 83.

Burgoyne TW, Hayes JM. 1998. Quantitative Production of H2 by Pyrolysis of Gas Chromatographic Effluents. Analytical Chemistry 70: 5136–5141.

Chikaraishi Y, Naraoka H. 2007. δ13C and δD relationships among three n-alkyl compound classes (n-alkanoic acid, n-alkane and n-alkanol) of terrestrial higher plants. Organic Geochemistry 38: 198–215.

Chikaraishi Y, Naraoka H, Poulson SR. 2004. Carbon and hydrogen isotopic fractionation during lipid biosynthesis in a higher plant (Cryptomeria japonica). Phytochemistry 65: 323–330.

Craig H, Gordon LI. 1965. Deuterium and oxygen 18 variations in the ocean and the marine atmosphere. Pisa, Italy: Consiglio Nationale Delle Ricerche Laboratorio Di Geologia Nucleare.

Dawson TE, Ehleringer JR. 1993. Gender-specific physiology, carbon isotope discrimination, and habitat distribution in boxelder, Acer negundo. Ecology: 798–815.

Dawson TE, Mambelli S, Plamboeck AH, Templer PH, Tu KP. 2002. Stable Isotopes In Plant Ecology. Annual Review of Ecology and Systematics 33: 507–559.

Eglinton G, Hamilton RJ. 1967. Leaf epicuticular waxes. Science 156: 1322–1335.

Eglinton TI, Eglinton G. 2008. Molecular proxies for paleoclimatology. Earth and Planetary Science Letters 275: 1–16.

Estep MF, Hoering TC. 1981. Biogeochemistry of the stable isotopes of hydrogen. Geochimica et Cosmochimica Acta 44: 1197-1206.

Feakins SJ, Sessions AL. 2010. Controls on the D/H ratios of plant leaf waxes in an arid ecosystem. Geochimica et Cosmochimica Acta 74: 2128–2141.

Gao L, Burnier A, Huang Y. 2011. Quantifying instantaneous regeneration rates of plant leaf waxes using stable hydrogen isotope labeling. Rapid Communications in Mass Spectrometry 26: 115–122.

Gao L, Tsai Y-J, Huang Y. 2012. Assessing the rate and timing of leaf wax regeneration in Fraxinus americana using stable hydrogen isotope labeling. Rapid Communications in Mass Spectrometry 26: 2241–2250.

Gat JR. 1996. Oxygen and hydrogen isotopes in the hydrologic cycle. Annual Review of Earth and Planetary Sciences 24: 225–262.

Helliker BR, Richter SL. 2008. Subtropical to boreal convergence of tree-leaf temperatures. Nature 454: 511–514. Chapter 1: General introduction 21

Hilkert AW, Douthitt CB, Schlüter HJ, Brand WA. 1999. Isotope ratio monitoring gas chromatography/Mass spectrometry of D/H by high temperature conversion isotope ratio mass spectrometry. Rapid Communications in Mass Spectrometry 13: 1226– 1230.

Huang Y, Shuman B, Wang Y, Webb T. 2004. Hydrogen isotope ratios of individual lipids in lake sediments as novel tracers of climatic and environmental change: a surface sediment test. Journal of Paleolimnology 31: 363–375.

Intergovernmental Panel on Climate I. 2014. Climate change 2014: impacts, adaptation, and vulnerability. Part A: global and sectoral aspects.

Jetter R, Schäffer S. 2001. Chemical composition of the Prunus laurocerasus leaf surface. Dynamic changes of the epicuticular wax film during leaf development. Plant Physiology 126: 1725–1737.

Jetter R, Kunst L, Samuels AL. 2006. Composition of plant cuticular waxes. In 'Biology of the plant cuticle (Eds M Riederer, C Müller) 145-181.

Kahmen A, Dawson TE, VIETH A, Sachse D. 2011a. Leaf wax n-alkane δD values are determined early in the ontogeny of Populus trichocarpa leaves when grown under controlled environmental conditions. Plant, Cell & Environment 34: 1639–1651.

Kahmen A, Hoffmann B, Schefuß E, Arndt SK, Cernusak LA, West JB, Sachse D. 2013a. Leaf water deuterium enrichment shapes leaf wax n-alkane δD values of angiosperm plants II: Observational evidence and global implications. Geochimica et Cosmochimica Acta 111: 50–63.

Kahmen A, Sachse D, Arndt SK. 2011b. Cellulose δ18O is an index of leaf-to-air vapor pressure difference (VPD) in tropical plants. Proceedings of the National Academy of Sciences of the United States of America 108: 1981-1986.

Kahmen A, Schefuß E, Sachse D. 2013b. Leaf water deuterium enrichment shapes leaf wax n- alkane dD values of angiosperm plants I: Experimental evidence and mechanistic insights. Geochimica et Cosmochimica Acta 111: 39–49.

Kahmen A, Simonin K, Tu KP, Merchant A, Cllister A, Siegwolf R, Dawson TE, Arndt SK. 2008. Effects of environmental parameters, leaf physiological properties and leaf water relations on leaf water δ 18O enrichment in different Eucalyptus species. Plant, Cell & Environment 31: 738–751.

Krull E, Sachse D, Mügler I, Thiele A. 2006. Compound-specific δ 13 C and δ 2 H analyses of plant and soil organic matter: a preliminary assessment of the effects of vegetation change on ecosystem hydrology. Soil Biology and Biochemistry 38: 3211–3221.

Leuzinger S, Luo Y, Beier C, Dieleman W, Vicca S, Körner C. 2011. Do global change experiments overestimate impacts on terrestrial ecosystems? Trends in Ecology & Evolution 26: 236–241.

McInerney FA, Helliker BR, Freeman KH. 2011. Hydrogen isotope ratios of leaf wax n-alkanes in grasses are insensitive to transpiration. Geochimica et Cosmochimica Acta 75: 541–554. 22 Chapter 1: General introduction

Pedentchouk N, Sumner W, Tipple B, Pagani M. 2008. δ 13 C and δD compositions of n- alkanes from modern angiosperms and conifers: an experimental set up in central Washington State, USA. Organic Geochemistry 39: 1066-1071.

Radke J, Bechtel A, Gaupp R, Püttmann W. 2005. Correlation between hydrogen isotope ratios of lipid biomarkers and sediment maturity. Geochimica et Cosmochimica Acta 69: 5517-5530.

Roden J, Ehleringer J. 1999. Observations of hydrogen and oxygen isotopes in leaf water confirm the craig-gordon model under wide-ranging environmental conditions. Plant Physiology 120: 1165–1174.

Roden JS, Lin G, Ehleringer JR. 2000. A mechanistic model for interpretation of hydrogen and oxygen isotope ratios in tree-ring cellulose. Geochimica et Cosmochimica Acta.

Sachs JP, Sachse D, Smittenberg RH, Zhang Z. 2009. Southward movement of the Pacific intertropical convergence zone AD 1400–1850. Nature Geoscience 2: 519-525.

Sachse D, Radke J, Gleixner G. 2004. Hydrogen isotope ratios of recent lacustrine sedimentary n-alkanes record modern climate variability. Geochimica et Cosmochimica Acta 68: 4877–4889.

Sauer PE, Eglinton TI, Hayes JM. 2001. Compound-specific D/H ratios of lipid biomarkers from sediments as a proxy for environmental and climatic conditions. Geochimica et Cosmochimica Acta 65: 213-222.

Schefuß E, Schouten S, Schneider RR. 2005. Climatic controls on central African hydrology during the past 20,000 years. Nature 437: 1003–1006.

Schimmelmann A, Sessions AL. 2006. Hydrogen isotopic (D/H) composition of organic matter during diagenesis and thermal maturation. Annual Review of Earth and Planetary Sciences. 34: 501–533.

Schmidt H-L, Werner RA, Eisenreich W. 2003. Systematics of 2H patterns in natural compounds and its importance for the elucidation of biosynthetic pathways. Phytochemistry Reviews 2: 61–85.

Sessions AL. 2006. Isotope-ratio detection for gas chromatography. Journal of Separation Science 29: 1946–1961.

Sessions AL, Burgoyne TW, Schimmelmann A. 1999. Fractionation of hydrogen isotopes in lipid biosynthesis. Organic Geochemistry 30: 1193-1200.

Sessions AL, Sylva SP, Summons RE. 2004. Isotopic exchange of carbon-bound hydrogen over geologic timescales. Geochimica et Cosmochimica Acta 68: 1545-1559.

Shepherd T, Wynne Griffiths D. 2006. The effects of stress on plant cuticular waxes. New Phytologist 171: 469–499.

Smith FA, Freeman KH. 2006. Influence of physiology and climate on δD of leaf wax n- alkanes from C3 and C4 grasses. Geochimica et Cosmochimica Acta 70: 1172–1187. Chapter 1: General introduction 23

Sternberg LDSL. 1988. D/H ratios of environmental water recorded by D/H ratios of plant lipids. Nature 333: 59–61.

Thompson LG, Mosley-Thompson E, Davis ME, Lin P-N, Henderson K, Mashiotta TA. 2003. Tropical Glacier and Ice Core Evidence of Climate Change on Annual to Millennial Time Scales. Advances in Global Change Research. Climate Variability and Change in High Elevation Regions: Past, Present & Future. Dordrecht: Springer Netherlands, 137–155.

Treydte KS, Schleser GH, Helle G, Frank DC, Winiger M, Haug GH, Esper J. 2006. The twentieth century was the wettest period in northern Pakistan over the past millennium. Nature 440: 1179–1182.

Zhang X, Gillespie AL, Sessions AL. 2009. Large D/H variations in bacterial lipids reflect central metabolic pathways. Proceedings of the National Academy of Sciences of the United States of America 106: 12580–12586.

24 Chapter 1: General introduction

Chapter 2: Drivers of leaf wax n-alkane δ2H values 25

Chapter 2

Effects of leaf water evaporative 2H-enrichment and biosynthetic fractionation on leaf wax n-alkane δ 2H values in

C3 and C4 grasses

Bruno Gamarra1,4, Dirk Sachse2,3 and Ansgar Kahmen1,4

1 Department of Environmental Systems Sciences, ETH Zürich, Switzerland

2 Institut für Erd- und Umweltwissenschaften, Universität Potsdam Germany

3 Helmholtz Centre Potsdam - GFZ German Research Centre for Geosciences, Section 5.1:

Geomorphology, Potsdam, Germany

4 Department of Environmental Sciences – Botany, University of Basel, Switzerland

Chapter 2 was published in the scientific journal Plant, Cell and Environment, September

2016.

26 Chapter 2: Drivers of leaf wax n-alkane δ2H values

2.1. Abstract

Leaf wax n-alkane δ2H values carry important information about environmental and ecophysiological processes in plants. However, the physiological and biochemical drivers that shape leaf wax n-alkane δ2H values are not completely understood. It is particularly unclear why n-alkanes in grasses are typically 2H-depleted compared to plants from other taxonomic groups such as dicotyledonous plants and why C3 grasses are 2H-depleted compared to C4 grasses. To resolve these uncertainties we quantified the effects of leaf water evaporative 2H-enrichment and biosynthetic hydrogen isotope fractionation on n- alkane δ2H values for a range of C3 and C4 grasses grown in climate-controlled chambers.

We found that only a fraction of leaf water evaporative 2H-enrichment is imprinted on the leaf wax n-alkane δ2H values in grasses. This is interesting, as previous studies have shown in dicotyledonous plants a nearly complete transfer of this 2H-enrichment to the n-alkane δ2H values. We thus infer that the typically observed 2H-depletion of n-alkanes in grasses (as opposed to dicots) is because only a fraction of the leaf water evaporative 2H-enrichment is imprinted on the δ2H values. Our experiments also show that differences in n-alkane δ2H values between C3 and C4 grasses are largely the result of systematic differences in biosynthetic fractionation between these two plant groups, which was on average -198‰ and -159‰ for C3 and C4 grasses, respectively.

Chapter 2: Drivers of leaf wax n-alkane δ2H values 27

2.2. Introduction

Stable isotope ratios of organic compounds synthesized by plants have been shown to correlate with environmental and plant physiological variables (Farquhar et al. 1982;

Dawson & Ehleringer 1993; Dawson et al. 2002; Barbour 2007; Craine et al. 2009). In particular the stable carbon, oxygen and nitrogen isotope composition of plant materials are now often used as proxies for climatic, environmental and physiological mechanisms in plants and ecosystems (Hobson 1999; Cerling et al. 1993; McLauchlan et al. 2013; Saurer et al. 2014). Recent technological developments now enable also the analysis of hydrogen isotope ratios (δ2H values) of plant-derived organic material (Filot et al. 2006). In particular the development of compound specific hydrogen isotope analyses has triggered increasing interest in applying hydrogen isotopes in environmental and physiological research. For example, this technology allows analyzing the hydrogen isotope composition of leaf wax lipids (Burgoyne & Hayes 1998; Hilkert et al. 1999; Sessions 2006), which are ideally suited to record and archive environmental and physiological information (Sachse et al. 2012).

Leaf wax lipids consist of long-chained hydrocarbons like n-alcohols, n-alkanoic acids and n-alkanes (Eglinton & Hamilton 1967) and are synthesized in the epidermal cells of plant leaves from where they are exported to the cuticle. The primary functions of the waxes are to preserve the plant from losing water, protect it against UV radiation and to act as a mechanical barrier against physical damage of the epidermis (Jetter et al. 2006). Leaf wax lipids, and in particular leaf wax n-alkanes, can persist in the environment for geological time scales. In addition, hydrogen in n-alkanes is mostly covalently bound to carbon so that the hydrogen isotope composition is remarkably stable over time and remains unchanged during early sedimentary diagenesis (Schimmelmann et al. 2006). Besides being highly persistent, n-alkanes are also ubiquitous in the environment and are abundantly found in soils, sediments and even in the atmosphere as aerosols (Eglinton & Eglinton 2008). n- 28 Chapter 2: Drivers of leaf wax n-alkane δ2H values

Alkanes and their δ2H values have therefore been successfully used in the past decade as biomarkers for paleoclimatic investigations (Sauer et al. 2001; Schefuß et al. 2005; Tierney et al. 2010; Sachse et al. 2012).

Current conceptual models suggest three major influences on leaf wax n-alkane δ2H values: (i) the plant’s source water δ2H values, (ii) the fractionation that occurs during leaf

2 water evaporation (εlw) leading to a H-enrichment of leaf water compared to the plant’s source water and (iii) the biosynthetic hydrogen isotope fractionation (εbio) that occurs during the biosynthesis of n-alkanes from leaf water and leads to a 2H-depletion of n-alkanes

(Schmidt et al. 2003; Sachse et al. 2004; Sachse et al. 2006; Smith & Freeman 2006; Zhang et al. 2009; Sachse et al. 2012; Kahmen et al. 2013a and b). Although the three main drivers that shape the δ2H values of leaf wax n-alkanes have now been identified, their interplay and relative contribution in driving variability in leaf wax n-alkane δ2H values across plant species, plant functional groups or across spatial and temporal gradients is poorly understood. This is in particularly true for n-alkanes in grasses that have generally been found to be 30‰ to 40‰ depleted in 2H when compared to n-alkanes from dicots (Liu &

Yang 2008; Sachse et al. 2012; Gao et al. 2014; Liu et al. 2015). Gao et al. (2014) argued that evolutionary differences in lipid biosynthesis are potential causes of different n-alkane δ2H values between grasses and dicot plants. They suggested that dicots possibly make use of

2H-enriched sugars stored in roots or other organs in the biosynthesis of leaf wax n-alkanes and that this is why leaf wax n-alkanes from grasses are 2H-depleted compared to those of

2 dicots. Likewise Liu et al. (2015) suggest that differences in εbio cause δ H values in wax n- alkanes of grasses being lower than those of dicots. However, robust and direct

experimental evidence for systematic differences in εbio between grasses and dicots does, to our knowledge, not exist. An alternative explanation for δ2H differences in leaf wax n- alkanes derived from grasses and dicots was provided by McInerney et al. (2011) and Chapter 2: Drivers of leaf wax n-alkane δ2H values 29

Kahmen et al. (2013a), who suggested that the effect of leaf water evaporative 2H- enrichment on n-alkane δ2H values differs between dicots and grasses. In a study under controlled environmental conditions Kahmen et al. (2013a) showed that dicots recorded

100% of the leaf water evaporative 2H-enrichment in their n-alkane δ2H values, while a C3 and a C4 grass recorded only a fraction of that signal (18% and 68%, respectively).

The different effects of leaf water evaporative 2H-enrichment on leaf wax n-alkane

δ2H values in dicots and grasses have been attributed to physiological and morphological differences between grasses and dicots (McInerney et al. 2011). Unlike dicots, grasses synthesize new leaf tissue in an intercalary meristem at the base of the leaf. A sheath protects this basal part of the leaf from losing water. Hence foliar water at the intercalary meristem, the primary location of leaf wax synthesis, is a mixture of 2H-unenriched source water and 2H-enriched water that diffuses downwards from the atmosphere-exposed leaf blade to the meristem. In contrast to dicots, leaf wax n-alkane δ2H values from grasses are therefore only partially affected by leaf water evaporative 2H-enrichment measured in the leaf blade, which could in turn explain why grasses are generally less 2H-enriched than dicots

(Kahmen et al. 2013a). Importantly, this explanation is based on very few experimental and observational data (McInerney et al. 2011; Kahmen et al. 2013a) and the general magnitude by which leaf water evaporative 2H-enrichment affects leaf wax n-alkane δ2H values in grasses remains yet to be quantified.

Within grasses systematic differences in n-alkane δ2H values have also been observed between C3 and C4 grasses. In general, C3 grasses have been reported to be 2H-depleted compared to C4 grasses (Smith & Freeman 2006; McInerney et al. 2011). Smith & Freeman

(2006) observed that C3 grasses had leaf wax n-alkane δ2H values on average 23‰ more 2H- depleted than C4 grasses for plants grown in greenhouses and of 21‰ for plants collected in the American Great Plains. They followed an earlier hypothesis from Helliker & Ehleringer 30 Chapter 2: Drivers of leaf wax n-alkane δ2H values

(2000) and explained δ2H differences between C3 and C4 grasses as a consequence of differences in leaf water evaporative 2H-enrichment between C3 and C4 plants (Helliker &

Ehleringer 2000; Smith & Freeman 2006; McInerney et al. 2011). In addition, McInerney et al. (2011) argued that differences in the source of organic substrates or use of different

NADPH pools in the different photosynthetic pathways could explain n-alkane δ2H differences between C3 and C4 plants. In the same study McInerney et al. (2011) show that

2 the apparent fractionation (“εapp”, i.e. the deviation of δ H values from source water to n- alkanes) varies from more negative to more positive values for C3 monocot, C4 monocot, C3 dicot and C4 dicot species, respectively indicating that life form also plays a major role in controlling the leaf wax n-alkane δ2H values. Similarly, Liu & Yang (2006) conclude that differences in n-alkane δ2H values among diverse plant life forms (e.g. tree, shrub, grass) are more significant than differences arising from a distinct photosynthetic pathway.

To separate the effects of source water δ2H values, leaf water evaporative 2H- enrichment and biosynthetic fractionation as drivers of leaf wax n-alkane δ2H differences in

C3 and C4 grasses a comprehensive study that allows the quantification of these effects across a range of different species will be necessary. In the study that we present here, it was therefore our objective to (i) quantify the extent by which leaf water evaporative 2H- enrichment shapes the δ2H values of leaf wax n-alkanes in different C3 and C4 grasses species, (ii) to determine if the consistently observed differences in leaf wax n-alkane δ2H values between C3 and C4 grasses can be explained by differences in leaf water evaporative

2H-enrichment and/or by different biosynthetic fractionation, and (iii) to establish if transpiration rates have an effect on the relationship between leaf water evaporative 2H- enrichment and leaf wax n-alkane δ2H values in C3 grasses. The overall motivation of our research was to improve the mechanistic basis that is necessary for the interpretation of Chapter 2: Drivers of leaf wax n-alkane δ2H values 31

grass-derived leaf wax n-alkane δ2H values in plant physiological and paleo-environmental research.

2.3. Material and Methods

2.3.1. Experimental design

In order to quantify the effect of leaf water evaporative 2H-enrichment on leaf wax n-alkane

δ2H values for C3 and C4 grasses and to evaluate if leaf water evaporative 2H- enrichment and/or biosynthetic fractionation determine the consistently observed differences in n-alkane δ2H values between C3 and C4 grasses, we conducted an experiment in climate-controlled growth chambers. For the experiment (first experiment) we used five

C3 grass species: Lolium perenne, Festuca rubra, Dactilys glomerata, Alopecurus pratensis,

Arrhenatherum elatius, and five C4 grass species: Panicum virgatum, Sporobolus cryptandrus, Sorghastrum nutans, Bouteloua curtipendula and Andropongon gerardii. To generate different leaf water δ2H values in our experimental plants we grew plants either in a chamber with a low relative humidity (dry treatment) or in a chamber with high relative humidity (wet treatment). Relative humidity in the dry treatment varied from 47.0% to

58.4% throughout the experiment and had a daily mean of 49.6%. Relative humidity in the wet treatment varied from 72.1% to 77.1% throughout the experiment and a daily mean of

74.3% (Figure 1). Other environmental variables such as air temperature and light intensity were held constant. Air temperature had values that ranged from 28.3°C to 29.0°C and a daily mean of 28.7°C in the dry treatment and from 28.1°C to 29.1°C and a daily mean of

28.7°C in the wet treatment. In both growth chambers light intensity during day had a constant value of 250 µmol/m2/s. Day/night cycles were 16h/8h.

For each treatment, we grew three replicates of each species from seeds in 20 x 20 cm

5-liter pots. To avoid evaporation of soil water, the surface of each pot was covered by 32 Chapter 2: Drivers of leaf wax n-alkane δ2H values

aluminum foil and a 1 cm layer of gravel (1 mm). Planting the seeds into the soil initiated the start of the experiment. All plants were irrigated with the same slightly 2H-enriched water, which had a δ2H value of 7.9‰. The slightly enriched water was obtained by mixing local tap water (-70‰) and highly enriched deuterium oxide (99.9 at%) in a 1000-liter tank. The isotopic composition of the irrigation water in the tank was monitored weekly and its isotopic composition remained constant (+/-0.2‰) throughout the experiment.

To determine the leaf water isotopic composition in C3 grasses, leaves were harvested in the wet and dry growth chambers 30, 43, 49 and 52 days after the beginning of the experiment from each replicate of the 5 species. For C4 grasses, leaf samples were collected from all three replicates of the five species on days 35, 51 and 62 after the beginning of the experiment. Leaf samples were stored in 10 ml exetainers at -20°C until water extraction. All leaf samples were taken during midday to capture maximum leaf water evaporative 2H- enrichment in a steady state. To determine source water isotope composition of the different species in the two treatments, soil water samples were taken from underneath the gravel layer of each pot in the wet and dry growth chambers 52 and 62 days after the beginning of the experiment for C3 and C4 plants respectively. As for the leaf samples, the soil samples were stored in 10 ml exetainers at -20°C until water extraction (see below). At the end of the experiment, which was day 52 for C3 grasses and day 62 for C4 grasses, leaves from all replicates were harvested, stored in paper bags and dried at 60°C. This leaf material was later used for lipid extractions (see below). To determine the isotopic composition of vapor in the growth chambers, air was trapped cryogenically 31, 35, 37, 42,

50 and 57 days after the beginning of the experiment. The air vapor cryogenic-traps consisted of 1m long tubes (5mm diameter) which were coiled 3 times and submerged in a mix of liquid nitrogen and ethanol at -90°C. A pump was connected at the other end of the Chapter 2: Drivers of leaf wax n-alkane δ2H values 33

tube and extracting air at a maximum flow rate of 10 liter/h, following the procedure described in Kahmen et al. (2008).

In a second experiment, we tested if high transpiration rates of a species can cause a particularly low impact of leaf water evaporative 2H-enrichment on leaf wax n-alkane δ2H values in grasses. This additional experiment was based on the hypothesis that a high transpiration stream would cause a large dilution of water at the intercalary meristem with

2H-unenriched source water (Helliker & Ehleringer 2002). To determine if the transpiration

2 rate of a species affected εlw and thus leaf wax δ H values, we selected seven C3 grass species: Lolium perenne, Festuca rubra, Dactilys glomerata, Alopecurus pratensis,

Arrhenatherum elatius, Holcus lanatus and Agrostis capillaris. We grew these species from seeds in 5-liter pots. Four replicates per species were grown in a growth chamber with a wet climate and four replicates per species were grown in a growth chamber with a dry climate.

The climate conditions in the wet and in the dry growth chambers as well as the isotope composition of the irrigation water were identical to the experiment (first experiment) described above. Leaves were sampled for leaf water and leaf wax n-alkane isotope analyses

37 days after the beginning of the experiment and water vapor was collected 8 and 28 days after the beginning of the experiment.

We measured the transpiration rates of all species 38 days after the beginning of the experiment. For this purpose, pots were watered in the night of day 37. At 5 am of the following day the pots were again watered to field capacity and the weight of each pot (w0) was recorded. The weight of each pot was again recorded 14 hours later (t1) immediately before the end of daylight in the chamber and again at 22 hours after last irrigation (t2) immediately before the lights went on again on day 38. We then harvested the whole remaining biomass and measured its dry weight (wbio) to express transpiration rates on a dry 34 Chapter 2: Drivers of leaf wax n-alkane δ2H values

weight basis (mmol kg-1 s-1). Daytime transpiration and night transpiration rates were then calculated as:

!! ! !! Equation 2.1.1: E!"# = !!"# × !!

!! ! !! Equation 2.1.2: E!"#$% = !!"# × !!

30 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● 28 ● 26 Figure 2.1. Air temperature

Air Temp (°C) 80 (°C), relative humidity (%), ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 2 70 source water δ H values 2 60 ● ● (‰, n=3), water vapor δ H RH (%) Wet chamber ● ● Dry chamber ● ● ● values (‰, n=2) and leaf ● ● ● ● ● ● ● ●●● ● ● ● ● 50 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● water δ2H values (‰, n=3) 40 20 ● Irrigation water in wet and dry treatments 10 ● ● ● ● ●

H (‰) (RH) during 65 days of the 2 0 first experiment. Wet and

Source water Source δ −80 ● ● ● dry treatments are −90 ● ● ● ● −100 represented by grey and ● ● H (‰) ● 2 −110 black colors, respectively. ● ● ● ● Water vapor δ −120 Wet chamber C3 and C4 grasses are ● ● Dry chamber −130 represented by triangles 40 and squares, respectively.

20 Mean values for the entire duration of the experiment 0 are shown on the right. The H (‰) 2

Leaf water water Leaf δ error bars denote 1SD of −20 C3 grasses wet chamber the mean. C4 grasses wet chamber C3 grasses dry chamber C4 grasses dry chamber −40 30 35 40 45 50 55 60 Dry Wet Days after the start experiment 1 Mean Chapter 2: Drivers of leaf wax n-alkane δ2H values 35

2.3.2. Water sample preparation and isotope analyses

Leaf and soil water from both experiments was cryogenically extracted as described by West et al. (2006). Briefly, we heated the sample containing exetainers in a water bath at 80°C.

The evaporated water escaping from the exetainers was trapped in U-tubes submerged into liquid nitrogen at -196°C. The exetainer U-tube system was under vacuum at 0.03 hPa. The water extraction process lasted 1.5 – 2 hours.

Hydrogen isotope analyses of soil water, leaf water and vapor were performed with a carbon reduction method following Gehre et al. (2004) at the Department of Environmental

Systems Sciences at ETH Zurich, Switzerland using a high-temperature elemental analyzer

(TC/EA) coupled to a DeltaplusXP isotope ratio mass spectrometer via a ConFlo III interface

(Finnigan MAT, Bremen, Germany). 0.5-1 µl of water was injected with a GC-PAL autosampler equipped with a gas tight syringe. Hydrogen and oxygen isotope ratios were determined using the peaks of H2 and CO. However we just report the hydrogen isotope values in this study. The position of the samples and standards in a measurement series, post-run offline calculations of offsets, memory effect and drift corrections to asses the final

δ2H values followed the concept developed by Werner & Brand (2001). All values were normalized to the V-SMOW standard. The internal laboratory standard had a δ2H value of -

80.53‰ during our measurements as compared to the target value of -80.27‰ with a standard deviation of 0.1‰.

2.3.3. Leaf wax sample preparation and isotope analyses

Grass leaves collected for n-alkane extractions were ground using a steel ball mill

(Rentsch). Total lipid extraction (TLE) was then done using an ultrasonic bath. For the ultrasonic bath approximatelly 1 g of ground leaf material was transferred into a 40 ml glass beaker. 30 ml of a mix of dichloromethane (DCM) and methanol (9:1) was 36 Chapter 2: Drivers of leaf wax n-alkane δ2H values

added and the beaker was placed into an ultrasonic bath for 15 minutes. The liquid phase contained the soluble lipids, such as alkyls (including n-alkanes), alcohols, fatty acids and esters whereas the solid remaining part had insoluble components of the plant material, e.g. polymers such as cellulose and lignine. The later phase was not considered for the rest of the analysis. Long-chained n-alkanes were purified from other apolar molecules with liquid chromatograpy (LC). For LC 6 ml glass columns were combustion cleaned in a combustion oven (500°C for 5 hours) and packed (¾ volume) with silica gel 60 (0.040-0.063 mm Alfa Aesar, Johnson Matthey Company)

99.5% pure. To obtain n-alkanes (fraction 1), the soluble lipids from the TLE were passed through the silica gel columns with 12 ml of n-hexane (GC-grade). To obtain alcohols (fraction 2), a mixture of hexane and DCM (1:1) was passed through the remaining column and to obtain the remaining apolar material (fraction 3), a mixture of DCM and methanol (9:1) was passed through the column. For the results reported here, only fraction 1 was considered.

n-Alkanes were identified and quantified using a gas chromatograph (Agilent

Technologies, 7890A) coupled with a flame ionization detector (FID) at the

Geologisches Institut, ETH Zurich. Measurements were done on a 30 m column

(Agilent, DB-5) with a diameter of 0.250 mm and film tickness of 0.25 µm. A subset of samples was measured at the Institut für Erd- und Umweltwissenschaften,

Universität Potsdam, and at the Department of Environmental Sciences - Botany,

University Basel. To quantify n-alkanes, peak area comparison was done with α- androstane, an internal laboratory standard. Total n-alkane concentrations are reported as µg n-alkane per g of dry leaf biomass. Chapter 2: Drivers of leaf wax n-alkane δ2H values 37

For the hydrogen isotope analyses of n-alkanes we used an isotope ratio mass spectrometer (Delta V Plus, ThermoFisher, Bremen, Germany) coupled to a GC

(Trace GC Ultra, ThermoFisher) via a ConFlow IV (ThermoFisher). n-Alkanes of different chain lengths were separated on a 30 m GC column (Agilent, DB-5) with a diameter of 0.250 mm and film tickness of 0.25 µm. The injections were done at a temperature of 270°C in splittless mode. For each analysis, we injected samples at a concentration of 300 ng/µl of the most abundant n-alkane dissolved in 1µL hexane.

Samples were analyzed in triplicates. The GC oven temperature was set to 90°C for 2 minutes, then it was raised at a rate of 10°C per minute to 150°C and finally at 4°C per minute to reach 320°C and then this temperature was held for 10 minutes. After separation on the GC column, individual n-alkane chain lenghts were pyrolized to H2 gas in an aluminum oxide reactor at 1420°C. In each sequence, a maximum 10 samples were injected in triplicates. An n-alkane standard laboratory mixture (A4, provided by A. Schimmelmann, Indiana University) was injected at three concentrations (100 ng/µl, 200 ng/µl, 400 ng/µl) at three times during each sequence. This resulted in 42 injections per sequence. Analysis of δ2H values from standards in different concentrations revealed that peak sizes below an area of 15 Vs produced unstable δ2H values. Therefore, n-alkanes with peak areas smaller that 15

Vs were omitted from the analyses. The linear relationship of known and measured

δ2H values from the A4 mixture was used to derive sample δ2H values relative to the

VSMOW scale. The precision of the IRMS was evaluated by the hydrogen isotopic composition stability of internal laboratory standard (nC29 alkane from Oak leaves) 38 Chapter 2: Drivers of leaf wax n-alkane δ2H values

which was analyzed three times in each sequence. Its mean value was -146.3‰ with a standard deviation of 2.8‰.

2.3.4. Data analysis

To compare leaf wax n-alkane δ2H values across treatments and species, we calculated the

2 concentration-weighed average (CWA) δ H values of nC29 and nC31, which were the most abundant compounds across the investigated species and allowed robust hydrogen isotope ratio measurements:

! ! ! (! !!!"# × !"#!.!!"# ! ! !!!"# × !"#!.!!"#) Equation 2.2: CWA δ H!!!"#!$% = !"#!.!!"# ! !"#!.!!"#

To quantify the effect by which leaf water evaporative 2H-enrichment affects the biosynthetic water pool and ultimately leaf wax n-alkane δ2H values in different C3 and C4 grass species we followed the approach of Kahmen et al. (2013a). For this purpose we

calculated for each species for both RH treatments the εlw, the fractionation that occurs during leaf water evaporation and leads to a 2H-enrichment of leaf water compared to soil

water. εlw is positive because the fractionation process discriminates against the light

(hydrogen) and favors the enrichment of the heavy isotope (deuterium):

! (! !!"#$ !"#$%!!) Equation 2.3: ε!" = ! − 1 (! !!"#$ !"#$%!!)

We then related εlw for each species and treatment to the apparent fractionation (εapp) which is the difference between n-alkane δ2H values and soil water δ2H values:

! ! !!!!"#!$%!! Equation 2.4: ε!"" = ! − 1 ! !!"#$ !"#$%!!

The slope of the relationship between εlw values and εapp indicates the magnitude by

2 which εlw affects the δ H values of leaf wax n-alkanes in a species, and can be expressed as the percentage when multiplied by 100. The intercept of this relationship resembles the

biosynthetic fractionation (εbio) that occurs during the biosynthesis of n-alkanes from the Chapter 2: Drivers of leaf wax n-alkane δ2H values 39

2 water pool in the leaf, leading to a H-depletion of n-alkanes. This estimate of εbio is based on the assumption that the humidity treatment in our experiment did not affect εbio in the investigated species. To test this assumption, we also employed an alternative approach to

estimate εbio using the following equation:

! ! !!!!"#!$%!! Equation 2.5: ε!"# = ! − 1 ! !!"!!

Stable isotope ratios are expressed in the delta (δ) notation and shown in per mil

(‰) as ratio of the heavy over the light isotope relative to the same ratio of an international standard according to the following equation:

! Equation 2.6: δ!H = 1000 × !"#$%& − 1 , ‰ !!"#$%#&%

To test for significant differences in leaf water d2H values, leaf wax n-alkane concentrations and leaf wax n-alkane d2H values across relative humidity (RH) treatments and, photosynthetic pathways (C3 and C4) we performed an analysis of variance (ANOVA) with RH treatment and photosynthetic pathway (C3 and C4) as factors for each of the two growth climate chambers. The statistical calculations were done using the statistical package

R version 3.1.1 (http://www.r-project.org).

2.4. Results

2.4.1. Leaf wax n-alkane concentrations in different humidity treatments

In the first experiment nC31 was the most abundant n-alkane in C3 and C4 grasses in the wet and the dry treatment (Table 2.1). The mean concentration of nC29, nC31 and nC33 alkanes across all C3 grasses was 45.6 µg/g, 60.8 µg/g and 23.6 µg/g in the wet treatment and 58.1

µg/g, 84.1 µg/g and 26.8 µg/g in the dry treatment. The mean concentration of nC29, nC31 and nC33 alkanes across all C4 grasses was 20.9 µg/g, 48.5 µg/g and 29.1 µg/g for the wet treatment and 50.7 µg/g, 95.4 µg/g, and 68.8 µg/g for the dry treatment, respectively. Total 40 Chapter 2: Drivers of leaf wax n-alkane δ2H values

leaf wax n-alkane concentrations, calculated as the sum of nC29, nC31 and nC33 alkanes, were statistically not different between C3 and C4 grasses (Table S2.1 in supporting information).

However, we found a marginally significant treatment effect where grasses in the dry treatment had on average higher total n-alkane concentration (168.9 µg/g for C3 and 214.9

µg/g for C4) than grasses in the wet treatment (130.0 µg/g for C3 and 98.6 µg/g for C4)

(Table 2.1 and Table S2.1 in supporting information). n-Alkanes with chain lengths nC25, nC27 and nC35 were not considered in this study since their concentrations were too low to allow robust hydrogen isotope analyses.

30 30

20 20

lw lw ε

ε 10 10 (‰) (‰) 0 0 −10 −10

H H 2 2

−140 −140 δ δ 258 17 25

−180 −180

-Alkane -Alkane −220 −220 (‰) n (‰) n

−140 258 17 25 17−140

app app −180 −180 (‰) ε ε (‰)

−220 −220

−140 258 17 258 17−140 bio bio ε −180 −180 (‰) ε

(‰) Wet chamber −220 Dry chamber −220

F. rubra A. elatiusMean C3 S. nutans Mean C4 L. perenne P. virgatum A. gerardii D. glomerataA. pratensis S. cryptandrusB. curtipendula

2 Figure 2.2. εlw (‰, n=4), n-alkane δ H values (‰, n=3), εapp (‰, n=3) and εbio (‰, n=3) of C3 and C4 grasses in the first experiment. Wet and dry treatments are represented by grey and black colors, respectively. Mean values of C3 and C4 grasses are shown on the right. The error bars denote 1SD of the mean. Statistical analyses of the data shown are presented in table 2. Chapter 2: Drivers of leaf wax n-alkane δ2H values 41

Table 2.1. Concentrations (µg/g) of leaf wax nC29, nC31 and nC33 and total n-alkanes of C3 and C4 grasses in the wet and dry treatment (RH) of the first and the second experiment.

RH nC29 nC31 nC33 Total

C3 grasses first experiment (µg/g) SD (µg/g) SD (µg/g) SD (µg/g) SD

L. perenne Wet 105.9 7.4 109.9 9.0 47.3 4.3 263.0 19.0 Dry 166.9 34.1 201.2 29.4 93.0 11.9 461.1 73.3 F. rubra Wet 81.0 8.0 165.2 17.3 21.8 3.3 268.1 24.8 Dry 88.8 13.8 190.2 29.9 5.4 5.1 284.4 39.2 D. glomerata Wet 13.0 1.8 0 0 13.0 1.8 Dry 8.4 1.9 3.5 1.1 4.0 2.3 15.9 5.2 A. pratensis Wet 11.8 6.5 11.4 0 15.6 12.2 Dry 12.2 0.5 7.8 1.8 2.0 18.1 5.5 A. elatius Wet 16.4 2.7 17.5 5.6 48.7 14.2 82.5 21.8 Dry 14.1 1.5 17.7 2.2 29.4 6.2 61.2 9.7 Mean Wet 45.6 60.8 23.3 130.0 SD 44.6 73.0 24.0 126.6 Mean Dry 58.1 84.1 26.8 168.9 SD 69.5 102.1 38.7 103.5

C4 grasses first experiment (µg/g) SD (µg/g) SD (µg/g) SD (µg/g) SD

P. virgatum Wet 34.0 20.9 48.3 30.2 3.8 2.0 86.7 52.9 Dry 140.2 12.0 138.2 7.6 12.0 4.3 290.5 7.6 S. cryptandrus Wet 23.3 7.6 127.7 53.0 118.4 42.9 269.3 101.6 Dry 36.2 8.6 235.1 107.3 297.6 172.2 568.9 285.5 S. nutans Wet 14.3 8.6 30.2 14.1 4.8 1.9 49.3 24.4 Dry 23.2 14.6 40.7 19.9 6.8 3.5 70.7 37.5 B. curtipendula Wet 24.4 18.4 25.9 18.8 13.9 10.3 64.2 47.5 Dry 39.5 24.4 39.8 14.9 18.7 7.1 98.0 46.1 A. gerardii Wet 8.4 1.0 10.5 1.3 4.5 0.7 23.4 2.4 Dry 14.2 3.3 23.1 8.5 8.9 4.9 46.2 16.4 Mean Wet 20.9 48.5 29.1 98.6 SD 9.9 46.3 50.1 98.2 Mean Dry 50.7 95.4 68.8 214.9 SD 51.1 90.4 128.0 220.2

C3 grasses second experiment (µg/g) SD (µg/g) SD (µg/g) SD (µg/g) SD

L. perenne Wet 69.8 11.6 103.1 12.7 72.3 13.0 245.3 36.6 Dry 86.6 27.5 120.3 27.5 85.5 14.0 292.4 68.1 F. rubra Wet 29.3 8.7 76.6 17.4 16.4 4.1 122.3 29.8 Dry 76.4 55.2 155.9 96.7 26.2 12.4 258.5 164.3 D. glomerata Wet 10.1 7.6 9.5 13.3 6.0 5.2 25.6 26.1 Dry 5.9 2.0 3.5 1.2 4.9 1.6 14.3 4.6 A. pratensis Wet 6.9 2.3 6.3 3.7 3.5 0.8 16.7 6.6 Dry 8.4 1.8 9.5 3.2 6.9 5.6 24.8 6.7 A. elatius Wet 12.3 2.7 12.0 4.3 28.8 8.7 53.0 14.6 Dry 9.2 2.2 23.1 8.5 40.1 6.6 72.3 12.1 H. lanatus Wet 3.7 1.1 1.2 -- 4.0 1.4 Dry 1.9 0.4 0.7 -- 2.1 0.6 A. capillaris Wet 5.7 1.8 5.0 1.6 5.6 1.6 16.3 4.9 Dry 13.8 2.4 14.8 3.7 9.8 3.0 38.3 8.6 Mean Wet 19.7 30.5 22.1 69.0 SD 23.7 41.4 26.4 87.4 Mean Dry 28.9 46.8 28.9 100.4 SD 36.2 63.6 30.8 122.0

42 Chapter 2: Drivers of leaf wax n-alkane δ2H values

2 2.4.2. Leaf water δ H values and εlw in C3 and C4 grasses

Water vapor collected in the wet and dry treatment throughout the experiment had mean

δ2H values of -112.5‰ (+/- 6.7‰) and -93.7‰ (+/- 6.8‰), respectively (Figure 2.1). For the

C3 plants the mean soil water δ2H values were -7.0‰ and -2.1‰ while for the C4 plants the mean soil water δ2H values were -5.8‰ and -4.5‰ for the wet and dry treatment, respectively. Soil water δ2H values were thus slightly 2H-enriched (5‰ - 10‰) in the dry treatment compared to the wet treatment (Table S2.2 in supporting information). This suggests that some soil evaporation did occur in the dry treatment although in our experiment we covered the surface of the pots with a layer of fine gravel to prevent it. We accounted for this effect by using soil water δ2H values (rather than irrigation water δ2H

2 values) to calculate εlw and εapp. The slight evaporative H-enrichment of soil water should thus not impact the outcome of our study.

In general, leaf water in all grasses was 2H-enriched in the dry compared to the wet treatment (Figure 2.1). In addition, leaf water δ2H values in C4 plants were on average 7.5‰ more 2H-enriched than that those of C3 grasses (Figure 2.1). This effect was statistically significant and independent of the humidity treatment (Table S2.1 in supporting information). Leaf water δ2H values varied throughout the first experiment by 7.0‰ in the wet and 5.2‰ in the dry treatment. These variations were small in comparison to the treatment differences. Across all C3 grasses the fractionation between leaf water and soil

water (i.e. εlw) showed a mean of -8.7‰ and 23.2‰ for the wet and the dry treatment, respectively, whereas C4 grasses had a mean εlw across all species of -4.6‰ and 23.3‰ for the wet and dry treatment, respectively (Figure 2.1). In general, differences in leaf water δ2H values and εlw were large between treatments but small (wet treatment) or absent (dry treatment) between C3 and C4 grasses (Table 2.2, Figure 2.2 and Table S2.2 in supporting information). Chapter 2: Drivers of leaf wax n-alkane δ2H values 43

Table 2.2. Results of ANOVAs testing the effects of the wet and dry treatment (RH) and

2 carbon fixation pathway (CFP) (i.e. C3 and C4) on εlw, n-alkane δ H values, εapp and εbio in the first experiment.

Df F value p-value LW Δ2H RH treatment 1 977.8 <0.001 Carbon fixation pathway (CFP) 1 5.0 0.028 RH treatment:CFP 1 6.1 0.016 groups (error) 74 n-Alkane δ2H RH treatment 1 21.8 <0.001 Carbon fixation pathway (CFP) 1 48.7 <0.001 RH treatment:CFP 1 0.0007 0.979 groups (error) 59

εapp RH treatment 1 13.8 <0.001 Carbon fixation pathway (CFP) 1 38.0 <0.001 RH treatment:CFP 1 0.2 0.628 groups (error) 64

εbio RH treatment 1 1.1 0.312 Carbon fixation pathway (CFP) 1 36.0 <0.001 RH treatment:CFP 1 0.0293 0.8648 groups (error) 64

2 2.4.3. Leaf wax n-alkane δ H values and εapp in C3 and C4 grasses

C3 grasses had mean leaf wax n-alkane δ2H values of -208.0‰ and -184.2‰ in the wet and in the dry treatment, respectively, while C4 grasses had mean leaf wax n-alkane δ2H values of -165.5‰ and -145.8‰ for the wet and dry treatment, respectively (Figure 2.2 and Table

S2.2 in supporting information). As such n-alkane δ2H values of all grasses were enriched in the dry compared to the wet treatment but the magnitude of this effect did not differ between C3 and C4 grasses (Table 2.2). Across all species n-alkane δ2H values in C4 grasses were 40‰ more enriched than n-alkane δ2H values in C3 grasses (Figure 2.2 and Table S2.2 in supporting information), which was independent of treatment (Table 2.2). To quantify the net 2H-fractionation between the soil water and leaf wax n-alkanes for C3 and C4 grasses we 44 Chapter 2: Drivers of leaf wax n-alkane δ2H values

2 calculated the apparent fractionation (εapp). In C3 grasses εapp for n-alkane δ H values was -

202.3‰ and -182.3‰ in the wet and dry treatment, respectively and for C4 grasses values

for εapp varied between -160.2‰ and -149.7‰ in the wet and dry treatment, respectively

(Table 2.2, Figure 2.2 and Table S2.2 in supporting information).

2 2.4.4. Effects of εlw and ε bio on the leaf wax n-alkane δ H values in C3 and C4 grasses

When we correlated εlw with εapp, we found positive relationships for both C3 and C4 grasses

(Figure 2.3). For different C3 grass species slopes ranged from 0.32 to 1.21 indicating that

2 2 32% to 121% of εlw and thus leaf water H-enrichment is recorded in the n-alkane δ H values.

For C4 species, the slopes ranged from 0.01 to 0.80 indicating that 1% to 80% of εlw and thus leaf water 2H-enrichment is recorded in the n-alkane δ2H values (Figure 2.3). On average, the

2 effect of εlw on n-alkane δ H values was stronger in C3 grasses (61%) than in C4 grasses

(39%). However, these differences were not statistically significant (Figure 2.4).

The intercepts of the linear relationship between εlw and εapp can be interpreted as biosynthetic fractionation between the biosynthetic (leaf-) water pool and the n-alkane

(Kahmen et al. 2013a). C3 grasses showed intercepts and thus εbio that ranged from -162‰ to -222‰ with an across species mean of -198‰ whereas in C4 grasses the intercepts and

thus εbio ranged from -144‰ to -175‰ with an across species mean of -159‰ (Figure 2.3).

Across species mean values for εbio differed significantly between C3 and C4 grasses (Figure

2.4). We used an additional approach to calculate εbio for individuals of each species in the

2 wet and in the dry treatment by subtracting εlw from n-alkane δ H values. The across species average values of εbio that we obtained were -196‰ and -202‰ for C3 and -157‰ and -

170‰ for C4 grasses in the wet and dry treatment, respectively (Table 2.3 and Figure 2.2).

Importantly, the εbio values obtained from the linear regressions were in the same range as

2 values that we obtained for εbio by subtracting the εlw from n-alkane δ H values (Table 2.3,

Figure 2.2 and 2.4). Chapter 2: Drivers of leaf wax n-alkane δ2H values 45

2 2 Table 2.3. Calculated εbio (‰, as the difference of n-alkane δ H values and leaf water δ H values) in the wet and dry treatment (RH) and estimated εbio (‰, as the intercept of the

2 relationship between εlw and n-alkane δ H values) of C3 and C4 grasses in the first and the second experiment.

Calculated Calculated Estimated

εbio εbio εbio (Wet) (Dry) Intercept

C3 grasses first experiment (‰) SD (‰) SD (‰) L. perenne -158.2 0.8 -173.6 3.2 -162.2 F. rubra -216.2 5.3 -229.0 2.4 -220.2 D. glomerata -193.6 2.8 -179.5 6.8 -189.6 A. pratensis -192.5 7.0 -200.8 0.6 -194.9 A. elatius -220.5 1.0 -224.6 2.4 -221.6 H. lanatus ------A. capillaris ------

Mean -196.2 -201.5 -197.7

SD 24.8 25.3 24.6

C4 grasses first experiment (‰) SD (‰) SD (‰) P. virgatum -173.8 5.0 -189.0 8.0 -174.9 S. cryptandrus -173.6 3.8 -189.6 3.2 -174.8 S. nutans -152.0 1.6 -159.6 7.2 -152.5 B. curtipendula -142.7 2.6 -145.3 8.8 -144.3 A. gerardii -144.5 -164.0 1.2 -149.2 Mean -157.3 -169.5 -159.1 SD 15.4 19.4 14.7

C3 grasses second experiment (‰) SD (‰) SD (‰) L. perenne -173.5 6.1 -189.7 3.0 -177.3 F. rubra -228.9 2.4 -243.8 2.4 -230.4 D. glomerata -173.1 4.1 -176.1 5.0 -173.4 A. pratensis -190.3 10.3 -204.4 8.7 -193.0 A. elatius -208.1 4.4 -212.2 3.9 -209.6 H. lanatus -177.9 3.9 -182.6 3.8 -178.8 A. capillaris -167.2 4.5 -190.3 1.6 -177.1 Mean -188.4 -199.9 -191.4 SD 22.5 22.9 21.4

2.4.5. Relation between transpiration and leaf wax n-alkane δ 2H values in C3 grasses

The environmental conditions in the growth chambers of the first experiment prevailed in the second experiment as described above. Water vapor had mean δ2H values of -105‰ and 46 Chapter 2: Drivers of leaf wax n-alkane δ2H values

-88‰ in wet and dry treatments, respectively. Leaf water δ2H values across all C3 grasses grown in the second experiment were on average -13.1‰ and 19.8‰ in the wet and dry

treatment, respectively. Accordingly εlw were -5.5‰ and 20.2‰ for the wet and dry treatment, respectively (Table S2.2 in supporting information). Total n-alkane concentrations were 69.0 µg/g and 100.4 µg/g for the wet and dry treatment, respectively

(Table 2.1). Grasses had mean leaf wax n-alkane δ2H values of -198.4‰ and -180.7‰ in the wet and dry treatment, respectively (Table S2.2 in supporting information). The relationships

between εlw and εapp indicated that, as already observed in the first experiment, εlw influenced leaf wax n-alkane δ2H values. Again, the effect was highly variable across species and slopes ranged between 0.09 and 0.67 with a mean value of 0.38 (Figure 2.5). Figure 2.5 shows the slopes of C3 grass species found in the first and in the second experiment. In both experiments, L. perenne, F. rubra and A. pratensis had the lowest values, lower than 0.5, and

A. elatius and D. glomerata had the highest slope values, higher than 0.6. εbio that we again estimated from the intercept of the relationship between εlw and εapp was -191‰ and thus similar to εbio for C3 grasses in the first experiment. Additionally we also estimated εbio in the

2 second experiment by subtracting εlw from n-alkane δ H values. The obtained values for εbio were -188‰ and -200‰ for the wet and the dry treatment, respectively (Table 2.3).

Transpiration rates were higher during the day than at night (Table S2.3 in supporting information). Additionally, transpiration in the dry treatment was substantially larger than in the wet treatment. Mean day transpiration rates were 42.9 mmol kg-1 s-1 and 66.1 mmol kg-1 s-1 and mean night transpiration rates were 19.4 mmol kg-1 s-1 and 23.7 mmol kg-1 s-1 for the wet and dry treatments, respectively. We observed substantial variability in transpiration rates across species as illustrated in Table S2.3 (in supporting information). We had hypothesized, that species showing a high transpiration rate will show a small effect of leaf water evaporative 2H-enrichment on leaf wax n-alkane δ2H values as the high transpiration Chapter 2: Drivers of leaf wax n-alkane δ2H values 47

rate dilutes the biosynthetic water at the intercalary meristem of leaves with 2H-unenriched source water. In contrast to our hypothesis, we found no significant relationship between

2 the effect of εlw on n-alkane δ H values and transpiration rates in the wet or the dry treatments during day and night periods (Figure 2.6).

−120 −120 ●

● ● −160 −160

(‰) (‰) app app ε ε −200 ● −200

−240 −240

−20 0 20 40 −20 0 20 40 ε ε lw (‰) lw (‰)

C3 grasses C4 grasses L. perenne P. virgatum y=0.32x−162 y=0.32x−175 F. r u b r a S. cryptandrus y=0.33x−220 y=0.29x−175 ● D. glomerata S. nutans y=1.21x−190 y=0.48x−152 ● A. pratensis B. curtipendula y=0.53x−195 y=0.80x−144 A. elatius A. gerardii y=0.66x−222 y=0.01x−149

Figure 2.3. Relationship between εlw (‰, n=3) and εapp (‰, n=3) of C3 and C4 grasses in the first experiment. Oblique grey lines (1:1) represent a theoretical 100% transfer of leaf water 2H-enrichment on n-alkane δ2H values.

48 Chapter 2: Drivers of leaf wax n-alkane δ2H values

app app ε ε 1.00 1.00 and and 0.75 0.75 lw lw ε ε 0.50 0.50

0.25 0.25 between between Slope of the relationship Slope of the relationship

−160 −160

−180 −180 (‰) (‰)

bio −200 −200 bio ε ε

−220 −220 * *

F. rubra A. elatiusMean C3 S. nutans Mean C4 L. perenne P. virgatum A. gerardii D. glomerataA. pratensis S. cryptandrusB. curtipendula

Figure 2.4. Slopes of the relationship between the εlw and εapp of C3 and C4 grasses (above) and (intercept-)estimated εbio (‰) (below) of C3 (left) and C4 (right) grasses in the first experiment. The error bars denote 1SD of the mean. “*” indicates significant differences between C3 and C4 grasses at = P < 0.05.

Chapter 2: Drivers of leaf wax n-alkane δ2H values 49

−150 (1:1) ● 1.0 −170 ● ● ● ●

(‰) −190 0.6 app ε −210 0.2 Slopes of second experiment −230 −20 0 20 40 0.2 0.6 1.0 ε lw (‰) Slopes of first L. perenne experiment y=0.26x−177 F. r u b r a A. elatius y=0.15x−230 y=0.67x−210 ● D. glomerata H. lanatus y=0.65x−173 y=0.63x−179 A. pratensis ● A. capillaris y=0.23x−193 y=0.09x−177

Figure 2.5. Relationship between the εlw (‰, n=4) and εapp (‰, n=4) of C3 grasses in the second experiment (left). Oblique grey lines (1:1) represent a theoretical 100% transfer of leaf water 2H-enrichement on n-alkane δ2H values. Correlation (right) between slopes of C3 grasses in first and second experiment.

2.5. Discussion

2.5.1. Leaf wax n-alkane concentrations in different humidity treatments

We found that n-alkane concentrations did not differ significantly between C3 and C4 plants but concentrations were higher in plants that were grown in the dry treatment compared to the wet treatment in both experiments (Table 2.1 and S2.1 in supporting information). The n-alkane concentrations we found are comparable to data found in previous studies for 50 Chapter 2: Drivers of leaf wax n-alkane δ2H values

grasses in natural environments (Zhang et al. 2004; Rommerskirchen et al. 2006; Vogts et al.

2009; Smith & Freeman 2006; McInerney et al. 2011; Gamarra & Kahmen 2015). Also, previous studies have observed that n-alkane concentrations in woody plants are higher under dry conditions (Sachse et al. 2006; Hoffman et al. 2013). Although the biological function of individual chemical compounds in the cuticle, among them n-alkanes, is still a matter of debate, it has been suggested that a high evaporative demand in the atmosphere triggers the synthesis of n-alkanes to increase the capacity of the cuticle to prevent evaporative water loss (Jetter & Schäffer 2001; Jetter et al. 2006).

2 2.5.2. Leaf water δ H values and εlw in C3 and C4 grasses

Leaf water δ2H values of C3 and C4 grasses were significantly 2H-enriched in the dry compared to the wet treatment (Figure 2.1). Leaf water being 2H-enriched in low humidity environments has been reported numerous times in the literature (Flanagan et al. 1991;

Roden & Ehleringer 1999; Barbour 2007) and is consistent with isotope theory (Craig &

Gordon 1965). Interestingly, C3 and C4 plants showed positive leaf water δ2H values in the dry treatment but negative values in the wet treatment. With a source water δ2H value of

+7.9‰ a “true” evaporative 2H-enrichment above soil water was therefore only observed in the dry treatment while leaf water δ2H values were 2H-depleted compared to soil water in the wet treatment. The reason for this is that in our study 2H-spiked irrigation water and atmospheric vapor were not in isotopic equilibrium, with vapor δ2H values that were much more negative than expected under equilibrium (Figure 2.1). This artifact was occasioned by introducing slightly 2H-enriched source water with the goal to maximize leaf water δ2H differences between the wet and the dry treatments (for details see Kahmen et al. 2013a).

At high humidity, water vapor exerts a strong control on the isotope composition of leaf water (Craig & Gordon 1965; Kahmen et al. 2008). In the wet treatment of our experiment, the negative water vapor δ2H values thus caused leaf water δ2H values to be 2H-depleted Chapter 2: Drivers of leaf wax n-alkane δ2H values 51

compared to the source water. Although it seems not intuitive to have leaf water δ2H values that are 2H-depleted compared to source water, this effect is in full agreement with isotope theory (Craig & Gordon 1965) and had no impact on the outcome of our study.

Leaf water was only slightly more 2H-enriched in C4 compared to C3 grasses (Figure

2.1). In the dry treatment, this effect can largely be attributed to differences in soil (source) water δ2H values, which we found to be 6.6‰ enriched in C4 compared to C3 plants. In fact,

2 across species average values for εlw (that accounts for soil water δ H values) were identical for C3 and C4 grasses in the dry treatment (Figure 2.2 and Table S2.2 in supporting information). In the wet treatment, different soil water δ2H values can also partially explain

2 different leaf water δ H values between C3 and C4 grasses so that across species average εlw values were only 4.1‰ more positive in C4 compared to C3 grasses. Helliker & Ehleringer

(2000) found that C4 grasses had a higher leaf water evaporative 18O-enrichment (14.7‰) than C3 grasses (7.2‰). They attributed these differences to the distinct differences in leaf anatomy between C3 and C4 grasses. C4 grasses tend to have shorter interveinal distances, which enhances the mixing of bulk leaf water with 2H-enriched lamina water. Additionally,

C3 and C4 grasses differ in their water use efficiency (WUE). C4 grasses are ecologically adapted to drier environments to economize water by having lower transpiration rates. In consideration of the Péclet effect (Dongmann et al. 1974; Farquhar & Lloyd 1993; Kahmen et al. 2011) this should result in higher leaf water evaporative 2H-enrichment in C4 than C3 grasses. Our data cannot confirm these previous findings and suggest that different leaf

anatomies and WUEs do not necessarily lead to differences in εlw between C3 and C4 grasses. Also this suggests that differences in εlw between C3 and C4 grasses cannot explain the different n-alkane δ2H values between C3 and C4 grasses that have been observed in our and previous studies (Helliker & Ehleringer 2000; Smith & Freeman 2006; McInerney et al.

2011; Sachse et al. 2012). 52 Chapter 2: Drivers of leaf wax n-alkane δ2H values

2 2.5.3. ε lw affects only partially the leaf wax n-alkane δ H values in C3 and C4 grasses n-Alkanes were 2H-enriched in the dry compared to the wet treatment in all C3 and C4 grass

2 species suggesting an influence of εlw on n-alkane δ H as previously proposed by Feakins &

Sessions (2010) and Kahmen et al. (2013a) (Figure 2.2 and Table S2.2 in supporting

information). We correlated εlw values and εapp to estimate the specific contribution of leaf water evaporative 2H-enrichment on n-alkane δ2H values in C3 and C4 species. Based on these relationships, we found that on average 61% (+/- 37%) and 38% (+/- 26%) of the leaf water evaporative 2H-enrichment was transferred to leaf wax n-alkane δ2H values in C3 plants in the first and second experiment, respectively and 39% (+/- 27%) in C4 plants in the first experiment. Our data thus confirm that only a fraction of leaf water evaporative 2H- enrichment is transferred to leaf wax n-alkane δ2H values in grasses as has been previously suggested (Kahmen et al. 2013a).

In Kahmen et al. (2013a) the effects of leaf water evaporative 2H-enrichment on the n-alkane δ2H values of grasses varied between 18% and 68% while the effect was larger

(96%) and highly consistent for different dicotyledonous species. The different magnitude by

2 which εlw affects the δ H values of leaf wax n-alkanes in grasses and dicotyledonous plants was explained by differences in leaf growth and development between these two plant groups (McInerney et al. 2011; Kahmen et al. 2013a). In grasses leaf development occurs at the intercalary meristem at the base of the leaf blade where a leaf sheath protects the new leaves from evaporative water loss. Leaf water at the intercalary meristem is a mix of 2H- unenriched xylem water and 2H-enriched leaf water that diffuses to the meristem from the evaporative sites in the leaf blade. In combination, these two water sources form the biosynthetic pool that the plant utilizes for the biosynthesis of leaf wax lipids and that is less

2H-enriched than water at the evaporative sites. Interestingly, Helliker & Ehleringer (2002) already suggested a mean effect of leaf water evaporative 18O-enrichment on cellulose δ18O Chapter 2: Drivers of leaf wax n-alkane δ2H values 53

values of 0.61 and 0.73 for C3 and C4 grasses, respectively. These findings agree with what we report here, that organic compounds in grasses do not record the full extent of heavy isotope enrichment of leaf water although the link between leaf water and the signal in the organic compound seems stronger for oxygen isotopes in cellulose than for hydrogen isotopes in lipids.

We observed a large variability in the slopes of the relationship between εlw and εapp for C3 grasses in both experiments and for C4 grasses in the first experiment. As such it is difficult to assign a specific fraction by which 2H-enriched leaf water is transferred to n- alkane δ2H values in grasses. We interpret the high across species variability in the slopes as the result of different δ2H values of the biosynthetic water pool at the primary sites of leaf wax formation at the intercalary meristem. In addition, we did not detect any significantly different slopes between C3 and C4 grasses (Figure 2.3 and 2.4). Based on this we conclude

2 that the effect of εlw on leaf wax n-alkane δ H values cannot explain the commonly observed differences between n-alkane δ2H values of C3 and C4 grasses.

2 3.5.4. ε bio as the main driver of different δ H values between C3 and C4 grasses

We observed significant differences in εbio between C3 and C4 grasses that were independent of the humidity treatment and the method that we used to calculate εbio (Table

2.2, 2.3, Figure 2.2, 2.4). εbio in C3 grasses was on average -198‰ while εbio in C4 grasses was on average -159‰. Importantly, differences in εbio between C3 and C4 grasses can thus explain the differences between leaf wax n-alkane δ2H values of C3 and C4 grasses that we identified in our study and that have been reported previously (Helliker & Ehleringer 2000;

Smith & Freeman 2006; McInerney et al. 2011). The significant differences in εbio between C3 and C4 grasses that we found here are also in agreement with observations of previous studies: Using modeled leaf water δ2H values Smith & Freeman (2006), for example, estimated εbio of -181‰ and -157‰ for C3 and C4 grasses, respectively. McInerney et al. 54 Chapter 2: Drivers of leaf wax n-alkane δ2H values

(2011) inferred εbio values of -198‰ and -175‰ for C3 and C4 grasses, respectively along a natural transect in the Great Plains of the US.

0.8 a (p = 0.9) b (p = 0.9) 0.8 ● ● 0.6 0.6 p 0.4 0.4 Figure 2.6. Relationship app app ε ε between slopes and 0.2 0.2 and ● ● and transpiration rates (mmol lw lw ε ε kg-1 s-1, n=4) of C3 grasses 0.8 c (p = 0.3) d (p = 0.4) 0.8 ● ● during day (a, c, e) and 0.6 0.6 night (b, d, f) in wet (a, b) p 0.4 0.4 and dry (c, d) treatments (RH) and the mean of both 0.2 0.2 ● ● treatments (e, f). None of the correlations revealed a 0.8 e (p = 0.8) f (p = 0.7) 0.8 ● ● significant relationship. 0.6 0.6 Slope of the relationship between p 0.4 0.4 Slope of the relationship between

0.2 0.2 ● ● 0 0 20 40 60 80 100 10 15 20 25 30 Transpiration Transpiration day (mmol/Kg/s) night (mmol/Kg/s) L. perenne A. elatius F. r u b r a H. lanatus ● D. glomerata ● A. capillaris A. pratensis

2 In the biosynthesis of n-alkanes εbio is determined by the δ H value of a carbohydrate precursor, the 2H additions from different NADPH sources, and the 2H transfer from surrounding water during biochemical reactions. Because the precursor in the biosynthetic

pathway of n-alkanes is the same for C3 and C4 plants, differences in εbio between C3 and C4 grasses should be either the result of different NADPH sources or the result of subsequent Chapter 2: Drivers of leaf wax n-alkane δ2H values 55

reactions with specific compartmented water in C3 and C4 plants. For example, the different anatomies between C3 and C4 leaves could influence the isotope exchange of hydrogen atoms of intermediate compounds with that of water in the mesophyll or bundle sheath cells, which might have different δ2H values. Although hydrogen transfer from compartmented-water pools can have an effect on εbio, such reactions contribute only 25% of the hydrogen atoms in lipid biosynthesis. On the other hand NADPH-derived hydrogen

sources account for 50% of the hydrogen in the εbio (the remaining 25% coming from the biochemical precursor molecule). Thus we deduce that differences in εbio between C3 and C4 grasses are most likely driven by the use of different NADPH sources during the biosynthesis of n-alkanes in C3 and C4 grasses. This is in particular since C4 plants, as compared to C3 plants, produce Malate-NADPH in the bundle sheath, which is more 2H-enriched than the

NADPH produced during the light reaction of photosynthesis (Smith et al. 1991).

2.5.5. The minor effect of transpiration on the relation between εlw and ε app

We hypothesized that high transpiration rates lead to a low contribution of leaf water evaporative 2H-enrichment to the biosynthetic water pool at the base of the leaves and thus a small influence of 2H-enriched leaf water on n-alkane δ2H values. We tested this hypothesis in our second experiment, where we correlated the slopes of the relationship

between εlw and εapp for different C3 grass species with the corresponding species daytime and nighttime transpiration rates. We, however, did not find a strong correlation between slopes and transpiration rates (Figure 2.6). We conclude therefore that transpiration rate alone cannot explain differences in the δ2H values of the biosynthetic water pool in the intercalary meristem across C3 grass species and that other anatomical or physiological factors contribute to the high variability in the slopes of different C3 grass species. One possible cause for this variability could be the secondary synthesis of leaf wax lipids after leaf maturation. n-Alkanes produced on mature leaf blades should have more positive δ2H 56 Chapter 2: Drivers of leaf wax n-alkane δ2H values

values as their biosynthesis would occur on leaf blades that are fully exposed to the

atmosphere and consequently also have a larger εlw. Variable rates of secondary n-alkane

2 synthesis could thus explain the variability of the effect of εlw on the leaf wax n-alkane δ H values observed in different species. Gao et al. (2011) suggest that secondary synthesis of lipids on fully matured grass blades of Phleum pratense is indeed high. However, Kahmen et al. (2011) for dicots and Gamarra & Kahmen (RCMS, in revision and in Chapter 3) for monocots could not confirm high secondary synthesis of leaf wax n-alkanes in fully matured grass leaves in seven different species, where secondary leaf wax n-alkane synthesis was no more than 1 percent per day. Such low rates of secondary leaf wax synthesis would have no substantial influence on the overall leaf wax signal. We thus propose that a mix of physiological (transpiration), anatomical (active/inactive water pools or mesophyll/xylem

ratios) and biochemical variables (NADPH source) influence the link between εlw and leaf wax n-alkane δ2H values but that no single variable can explain the observed variability across species.

2.5.6. Implications for the interpretation of leaf wax n-alkane δ 2H values derived from grasses

The work that we present here shows that leaf water evaporative 2H-enrichment is only partially reflected in n-alkane δ2H values of C3 and C4 grasses. The magnitude of this effect was on average 61% across all C3 and 39% across all C4 species with no significant differences between C3 and C4 grasses. Our data therefore suggest that more negative δ2H values in grasses as compared to previously reported δ2H values in dicotyledonous plants is caused by an incomplete transfer of εlw to the leaf wax n-alkanes. While we observed a large variability of this effect across different grass species, we were not able to identify a single physiological variable that could explain this variability. Our study also shows that systematic differences in n-alkane δ2H values between C3 and C4 grasses are not caused by differences in εlw between C3 and C4 grass species. However, we identified systematic differences in the Chapter 2: Drivers of leaf wax n-alkane δ2H values 57

biosynthetic fractionation of C3 and C4 plants. We thus conclude that it is εbio during lipid biosynthesis that drives the typically observed differences in n-alkane δ2H values between

C3 and C4 grass species. The mechanistic physiological information on what determines the n-alkane δ2H values in grasses has implications for the interpretation of n-alkane δ2H values in ecological and paleohydrological research. Our study confirms previous suggestions that biome shifts from woody vegetation to grassland and/or shifts from C3 and C4 vegetation can affect the sedimentary biomarker δ2H record beyond influences of the hydrological cycle. Our findings should be considered in the interpretation of the δ2H values of such records.

2.6. Acknowledgements

The authors would like to thank Tim Eglinton from the Geologisches Institut, ETH Zurich, for technical support and comments on the manuscript. We also thank Cesca McInerney,

Guillaume Tcherkez, Roland Werner and one anonymous referee for constructive comments on an early stage of the manuscript. BG and AK were funded by the ERC starting grant

COSIWAX (ERC-2011-StG Grant Agreement No. 279518). We disclose no conflict of interest in the development of this study.

58 Chapter 2: Drivers of leaf wax n-alkane δ2H values

2.7. Supplementary material

Table S2.1. Results of an ANOVA testing the effects of wet and dry treatment (RH) and carbon fixation pathway (CFP) (i.e. C3 and C4) on grass leaf wax n-alkane concentrations and leaf water δ2H values during the first experiment.

Df F value p-value n-Alkane concentration Relative humidity (RH) 1 3.8 0.063 Carbon fixation pathway (CFP) 1 0.03 0.862 RH:CFP 1 0.7 0.404 groups (error) 64 Leaf water δ2H values Relative humidity (RH) 1 21.8 <0.001 Carbon fixation pathway (CFP) 1 48.7 <0.001 RH:CFP 1 0.0007 0.979 groups (error) 74

Chapter 2: Drivers of leaf wax n-alkane δ2H values 59

2 2 Table S2.2. Soil water δ H values (‰), εlw (‰), leaf wax n-alkane δ H values (‰) and εapp (‰) of C3 and C4 grasses in the wet and dry treatment (RH) in the first and the second experiment.

Soil water n-Alkane 2 2 RH δ H εlw δ H εapp C3 grasses first experiment (‰) SD (‰) SD (‰) SD (‰) SD

L. perenne Wet -7.3 2.6 -7.6 2.4 -170.7 0.6 -164.6 2.8

Dry 0.3 2.5 22.3 2.9 -154.9 3.0 -155.1 1.5

F. rubra Wet -6.8 2.9 -8.8 3.3 -228.4 5.4 -223.1 6.1

Dry 0.0 2.5 19.8 1.5 -213.8 2.4 -213.8 3.4

D. glomerata Wet -5.6 0.9 -8.7 1.8 -202.1 2.3 -197.3 2.0

Dry -6.5 1.3 27.6 3.0 -160.9 5.2 -155.0 4.3

A. pratensis Wet -7.8 4.7 -8.0 4.5 -204.4 5.6 -198.4 8.5

Dry 1.6 3.7 18.2 3.2 -183.7 0.8 -184.6 3.0

A. elatius Wet -7.8 3.0 -10.2 4.5 -234.3 1.7 -228.3 3.8

Dry -5.8 1.0 28.2 2.7 -207.6 3.1 -203.0 3.5

H. lanatus Wet ------Dry ------A. capillaris Wet ------Dry

Mean Wet -7.1 -8.7 -208.0 -202.3

SD 0.9 1.0 25.2 25.3

Mean Dry -2.1 23.2 -184.2 -182.3

SD 3.8 4.5 26.6 27.0 C4 grasses first experiment (‰) SD (‰) SD (‰) SD (‰) SD

P. virgatum Wet -5.0 1.5 -3.1 1.4 -180.0 4.5 -175.4 4.9

Dry 4.8 0.7 28.9 0.8 -161.5 7.6 -165.5 8.0

S. cryptandrus Wet -6.9 2.6 -2.2 3.9 -181.1 2.2 -175.4 2.2

Dry -0.1 2.4 28.2 1.5 -165.3 2.4 -165.0 3.5

S. nutans Wet -8.1 1.9 -2.6 3.5 -158.9 1.2 -152.1 1.5 11. Dry 5.2 2.8 20.8 2.3 -137.5 8.0 -142.4 0

B. curtipendula Wet -5.1 1.5 -7.2 5.3 -155.0 1.1 -150.0 0.1

Dry 7.2 2.1 18.9 3.3 -122.9 8.6 -129.2 7.8

A. gerardii Wet -4.1 1.2 -7.8 1.7 -152.4 n.a -148.2 n.a

Dry 5.3 2.8 19.7 2.2 -141.8 0.8 -146.6 2.6

Mean Wet -5.8 -4.6 -165.5 -160.2

SD 1.6 2.7 14.0 13.9

Mean Dry 4.5 23.3 -145.8 -149.7

SD 2.7 4.9 17.6 15.6

60 Chapter 2: Drivers of leaf wax n-alkane δ2H values

C3 grasses second experiment (‰) SD (‰) SD (‰) SD (‰) SD

L. perenne Wet -8.3 1.1 -6.6 3.8 -185.8 2.7 -179.0 3.2

Dry -0.8 1.9 22.8 4.7 -171.9 2.5 -171.2 3.3

F. rubra Wet -6.7 0.7 -2.5 2.1 -236.0 1.3 -230.8 1.5

Dry 0.7 1.6 21.9 2.2 -226.7 1.9 -227.3 1.5

D. glomerata Wet -8.9 2.5 0.1 3.4 -180.7 1.4 -173.4 1.5

Dry 1.4 0.9 15.7 1.4 -161.9 5.4 -163.2 5.5 11. 11. A. pratensis Wet -8.5 2.4 -4.7 5.9 -200.9 7 -194.1 2

Dry -0.7 1.0 20.2 1.4 -188.8 9.4 -188.3 8.7

A. elatius Wet -7.1 0.8 -6.5 3.6 -219.5 3.3 -213.9 2.9

Dry -1.7 3.8 22.4 3.6 -195.9 4.3 -194.5 2.6

H. lanatus Wet -8.1 1.1 -4.6 1.2 -188.3 2.4 -181.6 3.0

Dry -1.6 2.9 19.9 3.5 -167.7 3.1 -166.4 4.5

A. capillaris Wet -6.1 2.1 -13.4 5.7 -183.3 5.0 -178.3 5.1

Dry 0.2 1.9 18.3 1.7 -175.7 1.4 -175.5 0.2

Mean Wet -7.7 -5.5 -198.4 -192.2

SD 1.0 4.2 19.6 20.1

Mean Dry -0.3 20.2 -180.7 -180.1

SD 1.2 2.5 22.6 23.1

Chapter 2: Drivers of leaf wax n-alkane δ2H values 61

Table S2.3. Transpiration rates (mmol kg-1 s-1) of C3 grasses in the wet and dry treatment (RH) and the mean (of both treatments) during the day (14 hours) and the night (8 hours) in the second experiment.

Transpiration RH Day SD Night SD (mmol/Kg/s) (mmol/Kg/s) L. perenne Wet 37.5 6.7 16.6 1.8 Dry 56.0 19.7 19.5 2.3 Mean 46.8 18.1 F rubra Wet 44.9 7.4 16.8 2.5 Dry 69.6 5.8 19.4 1.6 Mean 57.2 18.1 D glomerata Wet 55.4 10.8 20.6 2.3 Dry 89.7 15.4 29.7 4.1 Mean 72.6 25.1 A. pratensis Wet 31.6 4.3 20.7 2.1 Dry 52.2 17.1 25.4 7.4 Mean 41.9 23.1 A. elatius Wet 41.7 4.1 22.2 2.5 Dry 41.2 17.7 17.9 5.8 Mean 41.5 20.0 H. lanatus Wet 46.2 4.0 18.6 0.9 Dry 71.4 6.3 26.1 2.8 Mean 58.8 22.3 A.capillaris Wet 43.2 6.1 20.6 2.1 Dry 82.4 16.0 27.9 3.8 Mean 62.8 24.3 Mean Wet 42.9 19.4 Dry 66.1 23.7 Mean 54.5 21.6

62 Chapter 2: Drivers of leaf wax n-alkane δ2H values

2.8. References

Barbour M.M. 2007. Stable oxygen isotope composition of plant tissue: a review. Functional Plant Biology 34: 83-94.

Burgoyne T.W, Hayes J.M. 1998. Quantitative Production of H2 by Pyrolysis of Gas Chromatographic Effluents. Analytical Chemistry 70: 5136–5141.

Cerling T.E., Wang Y. & Quade J. 1993. Expansion of C4 ecosystems as an indicator of global ecological change in the late Miocene. Nature 367, 344-345.

Craig H., Gordon L.I. 1965. Deuterium and oxygen 18 variations in the ocean and the marine atmosphere. Pisa, Italy: Consiglio Nationale Delle Ricerche Laboratorio Di Geologia Nucleare.

Craine J.M., Elmore A.J., Aidar M.P.M., Bustamante M., Dawson T.E., Hobbie E.A., Kahmen A., Mack M.C., McLauchlan K.K., Michelsen A., Nardoto G.B., Pardo L.H., Peñuelas J., Reich P.B., Schuur E.A.G., Stock W.D., Templer P.H., Virginia R.A., Welker J.M. & Wright I.J. 2009. Global patterns of foliar nitrogen isotopes and their relationships with climate, mycorrhizal fungi, foliar nutrient concentrations, and nitrogen availability. New Phytologist 183, 980–992.

Dawson T.E., Ehleringer J.R. 1993. Gender-specific physiology, carbon isotope discrimination, and habitat distribution in boxelder, Acer negundo. Ecology: 798–815.

Dawson T.E., Mambelli S., Plamboeck A.H., Templer P.H., Tu K.P. 2002. Stable isotopes in plant ecology. Annual Review of Ecology and Systematics 33: 507–559.

Dongmann D.G., Nürnberg H.W., Förstel H. & Wagener K. 1974. On the enrichment of H2 18O in the leaves of transpiring plants. Radiation and Environmental Biophysics 11, 41–52.

Eglinton G., Hamilton R.J. 1967. Leaf epicuticular waxes. Science 156: 1322–1335.

Eglinton T.I., Eglinton G. 2008. Molecular proxies for paleoclimatology. Earth and Planetary Science Letters 275: 1–16.

Farquhar G.D., O'leary M.H., Berry J.A. 1982. On the relationship between carbon isotope discrimination and the intercellular carbon dioxide concentration in leaves. Functional Plant Biology 9: 121-37.

Farquhar G.D., Lloyd J. 1993. Carbon and oxygen isotope effects in the exchange of carbon dioxide between terrestrial plants and the atmosphere. In Stable Isotopes in Plant Carbon-Water Relations (eds J.R. Ehleringer, A.E. Hall & G.D. Farquhar), pp. 47–70. Academic Press, San Diego.

Feakins S.J., Sessions A.L. 2010. Controls on the D/H ratios of plant leaf waxes in an arid ecosystem. Geochimica et Cosmochimica Acta 74: 2128–2141. Chapter 2: Drivers of leaf wax n-alkane δ2H values 63

Filot M.S., Leuenberger M., Pazdur A. & Boettger T. 2006. Rapid online equilibration method to determine the D/H ratios of non-exchangeable hydrogen in cellulose. Rapid Communications in Mass Spectrometry 20, 3337–3344.

Flanagan L.B., Comstock J.P. & Ehleringer J.R. 1991. Comparison of Modeled and Observed Environmental Influences on the Stable Oxygen and Hydrogen Isotope Composition of Leaf Water in Phaseolus vulgaris L. Plant Physiology 96, 588–596.

Gamarra B., Kahmen A. 2015. Concentrations and δ²H values of cuticular n-alkanes vary significantly among plant organs, species and habitats in grasses from an alpine and a temperate European grassland. Oecologia 178: 981–998.

Gamarra B., Kahmen A. 2016. Variable and low secondary leaf wax n-alkane synthesis on fully mature leaves of C3 grasses. Rapid Communications in Mass Spectrometry (In review).

Gao L., Burnier A., Huang Y. 2011. Quantifying instantaneous regeneration rates of plant leaf waxes using stable hydrogen isotope labeling. Rapid Communications in Mass Spectrometry 26: 115–122.

Gao L., Edwards E.J., Zeng Y. & Huang Y. 2014. Major Evolutionary Trends in Hydrogen Isotope Fractionation of Vascular Plant Leaf Waxes (R Aroca, Ed.). PLoS ONE 9, e112610–10.

Gehre M., Geilmann H., Richter J., Werner R.A., Brand W.A. 2004. Continuous flow2H/1H and18O/16O analysis of water samples with dual inlet precision. Rapid Communications in Mass Spectrometry 18: 2650–2660.

Helliker B.R., Ehleringer J.R. 2000. Establishing a grassland signature in veins: 18O in the leaf water of C3 and C4 grasses. Proceedings of the National Academy of Sciences 97: 7894–7898.

Helliker B.R, Ehleringer J.R. 2002. Grass blades as tree rings: environmentally induced changes in the oxygen isotope ratio of cellulose along the length of grass blades. New Phytologist 155: 417–424.

Hilkert A.W., Douthitt C.B., Schlüter H.J., Brand WA. 1999. Isotope ratio monitoring gas chromatography/Mass spectrometry of D/H by high temperature conversion isotope ratio mass spectrometry. Rapid Communications in Mass Spectrometry 13: 1226– 1230.

Hobson K.A. 1999. Tracing origins and migration of wildlife using stable isotopes: a review. Oecologia 120, 314–326.

Hoffmann B., Kahmen A., Cernusak L.A., Arndt S.K., Sachse D. 2013. Abundance and distribution of leaf wax n-alkanes in leaves of Acacia and Eucalyptus trees along a strong humidity gradient in northern Australia. Organic Geochemistry 62: 62–67.

Jetter R., Schäffer S. 2001. Chemical composition of the Prunus laurocerasus leaf surface. Dynamic changes of the epicuticular wax film during leaf development. Plant Physiology 126, 1725–1737. 64 Chapter 2: Drivers of leaf wax n-alkane δ2H values

Jetter R., Kunst L., Samuels A.L. 2006. Composition of plant cuticular waxes. In: Riederer M, Müller C (eds) Biology of the plant cuticle. Blackwell, Oxford, pp 145–181

Kahmen A., Simonin K., Tu K.P., Merchant A., Callister A., Arndt S.K. 2008. Effects of environmental parameters, leaf physiological properties and leaf water relations on leaf water δ 18O enrichment in different Eucalyptus species. Plant, Cell & Environment 31: 738–751.

Kahmen A., Simonin K., Tu K., Goldsmith G.R. 2009. The influence of species and growing conditions on the 18-O enrichment of leaf water and its impact on ‘effective path length’. New Phytologist 184: 619-630.

Kahmen A., Dawson T.E., Vieth A., Sachse D. 2011. Leaf wax n-alkane δD values are determined early in the ontogeny of Populus trichocarpa leaves when grown under controlled environmental conditions. Plant, Cell & Environment 34: 1639–1651.

Kahmen A., Schefuß E., Sachse D. 2013a. Leaf water deuterium enrichment shapes leaf wax n-alkane dD values of angiosperm plants I: Experimental evidenceand mechanistic insights. Geochimica et Cosmochimica Acta 111: 39–49.

Kahmen A., Hoffmann B., Schefuß E., Arndt S.K., Cernusak L.A., West J.B. & Sachse D. 2013b. Leaf water deuterium enrichment shapes leaf wax n-alkane δD values of angiosperm plants II: Observational evidence and global implications. Geochimica et Cosmochimica Acta 111, 50–63.

Liu J., Liu W., An Z. 2015. An Insight into the reasons of leaf wax dD n- alkane values between grasses and woods. Sci Bull, 60: 549 555.

Liu W., Yang H. 2006. Hydrogen isotopic compositions of n-alkanes from terrestrial plants correlate with their ecological life forms. Oecologia 150: 330–338.

Liu W., Yang H. 2008. Multiple controls for the variability of hydrogen isotopic compositions in higher plant n-alkanes from modern ecosystems. Global Change Biology 14: 2166–2177.

McInerney F.A., Helliker B.R., Freeman K.H. 2011. Hydrogen isotope ratios of leaf wax n- alkanes in grasses are insensitive to transpiration. Geochimica et Cosmochimica Acta 75: 541–554.

McLauchlan K.K., Williams J.J., Craine J.M., Jeffers E.S. 2013. Changes in global nitrogen cycling during the Holocene epoch. Nature 495: 352–355.

Roden J., Ehleringer J. 1999. Observations of hydrogen and oxygen isotopes in leaf water confirm the craig-gordon model under wide-ranging environmental conditions. Plant Physiology 120: 1165–1174.

Rommerskirchen F., Plader A., Eglinton G., Chikaraishi Y., Rullkötter J. 2006. Chemotaxonomic significance of distribution and stable carbon isotopic composition of long-chain alkanes and alkan-1-ols in C4 grass waxes. Organic Geochemistry 37: 1303–1332. Chapter 2: Drivers of leaf wax n-alkane δ2H values 65

Sachse D., Radke J., Gleixner G. 2004. Hydrogen isotope ratios of recent lacustrine sedimentary n-alkanes record modern climate variability. Geochimica et Cosmochimica Acta 68: 4877–4889.

Sachse D., Radke J., Gleixner G. 2006. δD values of individual n-alkanes from terrestrial plants along a climatic gradient – Implications for the sedimentary biomarker record. Organic Geochemistry 37: 469–483.

Sachse D., Billault I., Bowen G.J., Chikaraishi Y., Dawson T.E., Kahmen A. 2012. Molecular Paleohydrology: Interpreting the Hydrogen-Isotopic Composition of Lipid Biomarkers from Photosynthesizing Organisms. Annu. Rev. Earth Planet. Sci. 40, 221–249.

Sauer P.E., Eglinton T.I., Hayes J.M., Schimmelmann A., Sessions A.L. 2001. Compound- specific D/H ratios of lipid biomarkers from sediments as a proxy for environmental and climatic conditions. Geochimica et Cosmochimica Acta 65: 213–222.

Saurer, M., Spahni, R., Frank, D.C., Joos, F., Leuenberger, M., Loader, N.J., McCarroll, D., Gagen, M., Poulter, B., Siegwolf, R.T.W., Andreu-Hayles, L., Boettger, T., Dorado Liñán, I., Fairchild, I.J., Friedrich, M., Gutierrez, E., Haupt, M., Hilasvuori, E., Heinrich, I., Helle, G., Grudd, H., Jalkanen, R., Levanič, T., Linderholm, H.W., Robertson, I., Sonninen, E., Treydte, K., Waterhouse, J.S., Woodley, E.J., Wynn, P.M. and Young, G.H.F. 2014. Spatial variability and temporal trends in water-use efficiency of European forests. Global Change Biology 20, 3700–3712.

Schefuß E., Schouten S. & Schneider R.R. 2005. Climatic controls on central African hydrology during the past 20,000 years. Nature 437, 1003–1006.

Schimmelmann A., Sessions A.L. & Mastalerz M. 2006. Hydrogen isotopic (D/H) composition of organic matter during diagenesis and thermal maturation. Annu. Rev. Earth Planet. Sci. 34, 501–533.

Schmidt H-L., Werner R.A., Eisenreich W. 2003. Systematics of 2H patterns in natural compounds and its importance for the elucidation of biosynthetic pathways. Phytochemistry Reviews 2: 61–85.

Sessions A.L. 2006. Isotope-ratio detection for gas chromatography. Journal of Separation Science 29: 1946–1961.

Smith B.N., Ziegler H. & Lipp J. 1991. Isotopic evidence for mesophyll reduction in Zea mays, an NADP-Malic enzyme plant. Naturwissenschaften 78, 358- 359.

Smith F.A., Freeman K.H. 2006. Influence of physiology and climate on δD of leaf wax n- alkanes from C3 and C4 grasses. Geochimica et Cosmochimica Acta 70: 1172–1187.

Tierney J.E., Oppo D.W., Rosenthal Y., Russell J.M., Linsley B.K. 2010. Coordinated hydrological regimes in the Indo-Pacific region during the past two millennia. Paleoceanography 25: PA1102.

Tipple B.J., Berke M.A., Hambach B., Roden J.S. & Ehleringer J.R. 2015. Predicting leaf wax n- alkane 2H/1H ratios: controlled water source and humidity experiments with hydroponically grown trees confirm predictions of Craig-Gordon model. Plant, Cell & Environment 38, 1035–1047. 66 Chapter 2: Drivers of leaf wax n-alkane δ2H values

Vogts A., Moossen H., Rommerskirchen F., Rullkötter J. 2009. Distribution patterns and stable carbon isotopic composition of alkanes and alkan-1-ols from plant waxes of African rain forest and savanna C3 species. Organic Geochemistry 40: 1037–1054.

Werner R.A., Brand W.A. 2001. Referencing strategies and techniques in stable isotope ratio analysis. Rapid Communications in Mass Spectrometry 15: 501–519.

West A.G., Patrickson S.J. & Ehleringer J.R. 2006. Water extraction times for plant and soil materials used in stable isotope analysis. Rapid Communications in Mass Spectrometry 20, 1317–1321.

Zhang Y., Togamura Y., Otsuki K. 2004. Study on the n-alkane patterns in some grasses and factors affecting the n-alkane patterns. The Journal of Agricultural Science 142: 469– 475.

Zhang X., Gillespie A.L., Sessions A.L. 2009. Large D/H variations in bacterial lipids reflect central metabolic pathways. Proceedings of the National Academy of Sciences of the USA 106: 12580–12586.

Chapter 3: Secondary n-alkane synthesis 67

Chapter 3

Low secondary leaf wax n-alkane synthesis on fully mature leaves of C3 grasses grown at low and high humidity

Bruno Gamarra12 and Ansgar Kahmen1,2

1 ETH Zurich, Institute of Agricultural Sciences, Zurich, Switzerland

2 UFZ, Helmholtz Centre for Environmental Research, Department of Community Ecology,

Halle, Germany

This chapter was submitted to the peer-reviewed scientific journal Rapid Communications in

Mass Spectrometry.

68 Chapter 3: Secondary n-alkane synthesis

3.1. Abstract

Leaf wax n-alkanes are long-chained aliphatic compounds that are present in the cuticle of terrestrial plant leaves. Their δ2H values are used for the reconstruction of past environments and for plant ecological investigations. The timing of n-alkane synthesis during the leaf development and the synthesis of secondary n-alkanes in fully matured leaves are still matter of debate. We estimated secondary leaf wax n-alkane synthesis in mature leaf blades of six C3 grass species grown in climate-controlled chambers. Secondary leaf wax n- alkane synthesis was determined by using a 2H-labeling approach. We found that mature grass leaves continue the synthesis of leaf wax n-alkanes after leave maturation. The rate of secondary n-alkanes synthesis was, however, relatively low and varied in response to atmospheric humidity and among species from 0.09 to 1.09% per day. Our investigation provides new evidence on the timing of cuticular wax synthesis in grass leaves and indicates that the majority of n-alkanes is synthesized during the development of the leaf at the intercalary meristem. Our study also sheds light on the interpretation of leaf wax n-alkane

δ2H values in environmental and geological studies as it suggests that secondary synthesis of leaf wax n-alkanes in grass’ leaves contributes only slightly to the geological record.

Chapter 3: Secondary n-alkane synthesis 69

3.2. Introduction

Leaf wax n-alkanes are long-chained aliphatic compounds that are synthesized in the epidermis of plant leaves and are then exported to the cuticle. The ecological function of n- alkanes – like other leaf waxes – is to protect the leaf from physical damage and the uncontrolled loss of water. n-Alkanes consist of covalently bounded C and H and lack any reactive groups and exchangeable hydrogen. Interestingly, n-alkanes and their hydrogen isotope composition (δ2H) can persist in the environment for millions of years. This is why leaf wax n-alkanes are establishing as powerful biomarker for the reconstruction of past environments (Sachse et al., 2012). In particular the δ2H values of leaf wax n-alkanes have been used as a robust tool for understanding past changes in the hydrologic cycle, plant- water relationships and more recently to study plant internal carbon metabolism (Schefuß et al., 2005; Smith and Freeman, 2006; Kahmen et al., 2013; Newberry et al., 2015; Tierney and

Russell). However, critical physiological/biochemical variables such as the influence of the leaf water on leaf wax n-alkane δ2H values or the timing of n-alkane synthesis during the development and life of a leaf are still not fully understood. This in turn hinders our interpretation of leaf wax n-alkane δ2H values in geological records or in plant physiological investigations.

Understanding the timing of leaf wax synthesis in the ontogeny and the subsequent life of a leaf is important. This is because the δ2H values of leaf n-alkanes could be shaped under the influence of environmental and leaf physiological processes occurring only during the short developmental phases of a leaf or over the entire lifespan of a leaf. In a recent review paper, Sachse et al. (2012) have stressed that across species difference in the timing and duration of leaf wax synthesis might be the cause of great δ2H variability observed in natural ecosystems. 70 Chapter 3: Secondary n-alkane synthesis

Today there is still no consensus if and when the production of leaf waxes ends during leaf ontogeny and if and to what degree fully matured leaves are able to synthesize leaf waxes. Early studies addressing this question have monitored the change in leaf wax composition and abundance over the lifespan of a leaf. Hauke and Schreiber (1998) for example found a change in wax composition during leaf development followed by a period of stable leaf wax abundance and composition after leaf maturation in Hedera helix L.

Additionally they detected a decrease in total wax amount at the last part of the season.

Jetter and Schäffer (2001) studied the abundance and composition of intra and epicuticlar waxes in leaves of Prunus laurocerasus. They found that leaf waxes were synthesized especially during the first stages of leaf development and that leaf wax productivity ceased later in the life of a leaf. Based on these results they propose an absence of spontaneous leaf way synthesis on fully mature leaves but suggest the possibility of replacement of cuticular waxes especially under environmental stress.

Pedentchouk et al. (2008) investigated the seasonal variability of δ2H values at natural abundance in leaf waxes and suggested continuous leaf wax synthesis and replacement of the cuticle, in particular under natural stress conditions. Also Sachse et al. (2009) observed seasonal changes of leaf wax δ2H values in Acer pseudoplatanus and Fagus sylvatica leaves over relatively short periods. They interpreted these changes as de novo leaf wax synthesis in mature leaves. In contrast, n-alkane δ2H values of barley grass leaves growing under natural conditions were found to change insignificantly throughout a leafs lifespan (Sachse et al., 2010). The authors thus suggest that δ2H values of grass leaves record precipitation or physiological processes only during leaf emergence. This is in agreement with results from

Feakins and Sessions (2010) who observed no seasonal variability on n-alkane δ2H values in the oak Quercus agrifolia during 10 months sample campaign suggesting also exclusive early synthesis of lipid waxes in the life of a plant leaf. Also Sachse et al. (2015) concluded that de Chapter 3: Secondary n-alkane synthesis 71

novo lipid synthesis only occurs during the first weeks of leaf expansion in Q. agrifolia. In the broadleaf Populus angustifolia. Tipple et al. (2013) also found that leaf waxes δ2H values only record the environmental condition during leaf flush at the beginning of the growing season.

It is important to note, however, that the natural variability of leaf wax abundances and δ2H values cannot inform about the de-novo synthesis of leaf waxes with certainty. This is because changes in leaf wax abundance and/or their δ2H values could be caused by de- novo synthesis but also by losses of compounds from the cuticle. Likewise consistent leaf wax abundances and δ2H values throughout the life of a leaf could either be the result of a ceased production or because losses and de-novo synthesis are exactly balanced. For this reason, several recent experiments have employed stable isotope tracers in the form of 2H- enriched water to directly study the temporal dynamics of cuticular n-alkanes. Using such an approach Kahmen et al. (2011) showed that 99 per mil 2H-enriched water applied as irrigation water was detected in the leaf water of Populus tricocarpa leaves, irrespective of their developmental stage. However, the 2H signal was incorporated only into n-alkanes of leaves that developed during the time of tracer application but not into n-alkanes of leaves that had already fully matured before tracer application. The authors concluded that insignificant amount of n-alkanes are synthesized in P. thricocarpa after leaf maturation. Gao et al. (2012) also investigated in a labeling study the lipid biosynthesis rates in Fraxinus americana. They also found a very low de-novo lipid synthesis in mature leaves and corroborated the earlier results by Kahmen et al. (2011) In contrast, Gao et al. (2011) described large de-novo n-alkane synthesis in matured leaves of the grass Phleum pratense.

2 They applied a tracer of 104 000 per mil and estimated based on δ H analyses in C27 n- alkanes a 1% de-novo synthesis per day in mature leaves. The limited number of tracer studies and the conflicting results call for additional work that tests and quantifies the ability 72 Chapter 3: Secondary n-alkane synthesis

of a broad range of different plant species to synthesize leaf waxes after their leaves have fully matured.

Here we tested if and to what extent secondary leaf wax synthesis occurs in mature leaf blades across a range of different C3 grass species by determining the incorporation of highly 2H-enriched irrigation water into leaf wax n-alkane δ2H values of mature grass leaves.

Differently than in dicotyledonous plants, whose leaves grow and expand after unfolding, grass leaves grow and synthesize leaf tissue at the intercalary meristem at the base of the leaf. There is a sheath developed from the previous old leaf that covers the leaf basal area and protects the new leaf from physical damage and transpirational water loss. The primary synthesis of cuticular lipids, including n-alkanes, occurs at the intercalary meristem.

Secondary leaf wax synthesis on mature grass leaves that are exposed to the atmosphere and thus subject to leaf water 2H-enrichment would have a strong impact of total leaf wax n- alkane δ2H values and the environmental signals recorded therein.

3.3. Material and Methods

3.3.1. Experimental setup

To determine the existence of secondary leaf wax n-alkane synthesis after leaf emergence and maturation in grasses, we grew six C3 grass species, Lolium perenne, Arrhenatherum elatius, Dactylis glomerata, Holcus lanatus, Anthoxanthum odoratum and Festuca rubra in growth chambers under controlled environmental conditions. The climate in the chambers was either wet with a RH of ∼75% or dry with a RH of ∼50%. The air temperature was held constant at 28.0°C and 28.2°C in the wet and dry climate growth chambers, respectively.

Light intensity had constant values of 250 µmol/m2/s in both chambers with the diurnal light cycle 16 hours day vs. 8 hours night. Plants were grown from seed in 20 cm x 20 cm 5 liter pots. Each species was replicated four times. Chapter 3: Secondary n-alkane synthesis 73

Table 3.1. Leaf wax concentrations of nC29, nC31, nC33 and total n-alkane (µg/g) of C3 grasses under wet and dry RH at t1 and t2.

t1 Total RH nC29 SD nC31 SD nC33 SD n-alkane (µg/g) (µg/g) (µg/g) (µg/g) L. perenne Wet n.a. n.a. n.a. n.a. Dry 106.7 23.3 190.1 35.9 188.2 53.1 485.0 A. elatius Wet 34.9 23.2 62.8 42.6 162.9 114.7 260.7 Dry n.a. n.a. n.a. n.a. D. glomerata Wet 25.3 22.5 23.1 ------48.4 Dry 32.3 37.8 16.4 19.9 12.3 13.7 61.0 H. lanatus Wet 10.3 3.2 4.4 1.0 -- -- 14.6 Dry n.a. n.a. n.a. n.a. A. odoratum Wet 55.5 25.4 96.1 42.8 156.7 74.0 308.3 Dry 72.2 24.8 141.4 58.6 166.3 11.4 379.9 F. rubra Wet 74.2 14.3 179.9 43.8 25.8 7.0 279.9 Dry 189.1 104.7 424.3 231.7 74.4 35.7 687.7 Species Mean Wet 40.1 73.3 115.1 182.4 SD Wet 25.2 69.4 77.4 139.3 Species Mean Dry 100.1 193.0 110.3 403.4 SD Dry 66.7 170.6 81.9 261.6

t 2 Total RH nC29 SD nC31 SD nC33 SD n-alkane (µg/g) (µg/g) (µg/g) (µg/g) L. perenne Wet n.a. n.a. n.a. n.a. Dry 24.8 1.9 40.6 1.7 38.0 1.7 103.5 A. elatius Wet 17.3 2.9 34.7 9.0 69.0 17.8 121.0 Dry n.a. n.a. n.a. n.a. D. glomerata Wet 32.6 20.6 30.9 -- -- 63.5 Dry 18.4 6.4 9.2 1.1 5.1 2.5 32.7 H. lanatus Wet 6.8 0.9 4.4 0.3 -- 11.1 Dry n.a. n.a. n.a. n.a. A. odoratum Wet 37.9 17.1 70.4 37.0 110.7 60.7 219.0 Dry 55.8 49.4 99.9 70.8 163.8 120.0 319.4 F. rubra Wet 58.1 16.0 132.1 46.5 18.4 2.6 208.7 Dry 64.4 37.2 182.6 38.1 38.2 6.7 285.1 Species Mean Wet 30.5 54.5 66.1 124.7 SD Wet 19.7 49.3 46.2 129.6 Species Mean Dry 40.8 83.1 61.3 185.2 SD Dry 22.6 76.2 70.1 138.9

74 Chapter 3: Secondary n-alkane synthesis

From day 1 to day 61 we used irrigation water with a δ2H value of -40‰. On day 61 we harvested a subset (30%) of leaves as our reference samples that we call from here on

2 ‘t1’ to determine leaf water and n-alkane δ H values. We then changed the isotopic composition of the irrigation water in both chambers to highly 2H-enriched water with a δ2H value of 1000‰. At the same time we gently marked the base of the grass leaves that were still attached to the plants with a permanent white marker. We did this to be able to distinguish leaf blades that had emerged and were already mature before we started our tracer application from leaf blades that emerged during tracer application. On day 120 and

123 of the experiment, i.e. after 59 and 62 days of consecutive irrigation with 2H-enriched water, we harvested the previously marked leaves by cutting the leaf blades above the white mark. For our latter analyses we term these leaf samples ‘t2’. Leaves harvested on day 120

2 were used to determine the leaf water δ H values of plants at t2 and leaves harvested on day

123 were used to determine n-alkane δ2H values. Leaves from different individuals growing in one pot were bulked into a single sample. Leaf samples were then stored in paper bags and dried at 60°C. Leaf samples were collected for leaf water δ2H values were stored in 10 ml sealed glass exetainers at -10°C until water extraction (see below).

At t2 three species (L. perenne in the wet chamber and A. elatius, H. lanatus in the dry chamber) turned to yellowish color for a reason we cannot explain. We did not harvest these plants and did not use them for the rest of the analysis.

3.3.2. Leaf water extraction and analyses

Leaf water was extracted from all samples with a cryogenic extraction line. The exetainers containing the samples were heated in a water bath at 80°C. The evaporated water was trapped in glass U-tubes that were submerged in liquid nitrogen at -196°C. The system was bombed for 2 hours with a vacuum of 0.03 hPa (West et al., 2005). Chapter 3: Secondary n-alkane synthesis 75

Hydrogen isotope composition of leaf water was analyzed with a carbon reduction method (Gehre et al., 2004) using a high-temperature elemental analyzer (TC/EA) coupled to a DeltaplusXP isotope ratio mass spectrometer via a ConFlo III interface (Finnigan MAT) at the

Department of Environmental Systems Sciences at ETH Zurich. 1 µL of water was injected with a GC-PAL autosampler equipped with a gas tight syringe. Deuterium and 18-oxygen

18 values were calculated using peaks of H2 and CO (δ O values are not considered in this study). The position of samples and lab standards in a sequence, post-run offline calculations of offsets, memory effects and drifts to calculate sample δ2H values which were normalized to the V-SMOW standard. All these concepts followed Werner and Brand (2001). During this study the internal laboratory standard had a δ2H value of -78.7‰ as compared to the target value of -77.9‰ with a standard deviation of 0.76.

3.3.3. n-Alkane extractions and analyses

To purify the wax lipids from the plant material, leaves were dried at 60°C for 24 hours.

About 1 g of dried leaf material was then clipped into small pieces for total lipid extraction

(TLE). The clipped leaves were placed in 40 ml beakers. TLE was obtained by transferring the beakers, with a mixture of dichloromethane (DCM) and methanol (MeOH) (30 ml; 9:1) in an ultrasonic bath for 15 minutes (Peters et al., 2005; Christie and Han, 2010). The aim is to isolate in a solution the apolar compounds of the leaf waxes, which include alkanes, fatty acids, alcohols and esters from the polar components. To separate n-alkanes from other apolar compounds we prepared 6 ml glass columns (columns were combusted at 500°C for 5 hours to remove organic residue) for liquid chromatography (LC). Three quarters of the column volume was packed with silica gel 60 (0.040-0.063 mm, 99.5% pure, Alfa Aesar,

Johnson Matthey Company) and rinsed with 2 volumes of acetone, 2 volumes DCM and 2 volumes of hexane. The packed columns were dried in a desiccation oven at 60°C for 12 hours for their chemical activation. n-Alkanes were separated by pipetting the TLE onto the 76 Chapter 3: Secondary n-alkane synthesis

packed columns with 12 ml of n-hexane (GC-grade) which resulted in ‘fraction 1’. A mixture of hexane and DCM (12 mL; 1:1) and DCM and MeOH (12 mL; 9:1) was repeatedly pipetted on the column thereafter, to obtain the fatty acids and the remaining apolar compounds which constituted ‘fraction 2’ and ‘fraction 3’, respectively (fractions 2 and 3 are not considered in this study).

n-Alkanes were identified and quantified in the fraction 1 using a gas chromatograph

(GC; Agilent Technologies, 7890A) coupled with a flame ionization detector (FID).

Measurements were done on a 30 m column (Agilent, DB-5) with a diameter of 0.250 mm and film thickness of 0.25 mm. n-Alkane quantification was done via peak area comparison with an internal standard (α-androstane) and concentrations are reported as n-alkane in µg over leaf dry biomass in g (µg/g).

An isotope ratio mass spectrometer (IRMS; Delta V Plus, ThermoFisher) coupled to a

GC (Trace GC Ultra, ThermoFisher) was used to measure the hydrogen isotope composition of the leaf wax derived n-alkanes. Samples were dissolved in hexane with a concentration of around 300 ng of the most abundant n-alkane per 1µL hexane. 1µL of the hexane solution was injected for an analysis. Separation of n-alkanes with different chain lengths was performed in a GC with a 30 m column (Agilent, DB-5) with a diameter of 0.250 mm and film thickness of 0.25 mm. The injector was in splittless mode at a temperature of 270°C. After injection the temperature of the GC oven was kept at 90°C for 2 minutes, then rose to 150°C at a rate of 10°C per minute, and to 320°C at a rate of 4°C per minute. This final temperature was held for 10 minutes. After compound separation on the GC column, individual alkanes were pyrolyzed to H2 gas in an aluminum oxide reactor at 1420°C. For each sequence, 10 samples were run in triplicates. Additionally, an n-alkane standard mixture (A4, provided by

A. Schimmelmann, Indiana University) at low, medium and high concentrations (100 ng/µl,

200 ng/µl, 400 ng/µl respectively) and an internal laboratory standard (nC29 alkane from Chapter 3: Secondary n-alkane synthesis 77

Californian Oak leaves) were each run at three different times during each sequence. The sequences thus consisted on a total of 42 injections. Analysis of the A4 mixture revealed that intensity peak sizes below areas of 15 Vs produced unstable isotope measurements.

Therefore, samples with peak sizes smaller than 15 Vs were excluded from further analyses.

The linear relationship of known and measured hydrogen isotope values of the A4 mixture was used to calculate sample δ2H values relative to the VSMOW scale. Analysis of δ2H values from the internal laboratory standard revealed in this study an instrumental precision of

0.9‰ measured as mean standard deviation. The H3+ factor, calculated before each sequence, had a mean value of 3.0.

3.3.4. Data analysis

In this study, we report n-alkane concentration values of nC29, nC31, nC33 and the total n- alkane concentrations. Total n-alkane concentrations were calculated as the sum of nC29, nC31 and nC31. To report hydrogen isotope ratios we determined the concentration weighed

2 average (CWA) δ H values of the most abundant n-alkanes (nC29, nC31 and nC33) of the investigated grass species:

! ! ! ! (! !!"!" × !"#!.!"!" ! ! !!"!" × !"#!.!"!"! ! !!"!!× !"#!.!"!!) Equation 3.1: CWA � �!!!"#!$% = !"#!.!"!" ! !"#!.!"!"! !"#!.!"!!

To quantify secondary synthesis of leaf wax n-alkanes in mature grass leaves, we followed the method developed by Gao et al. (2011; 2012). In brief, the fraction of n-alkanes

2 newly synthesized between t1 and t2 (f) can be calculated with δ H values of n-alkanes at t2

2 and the background n-alkane δ H values at t1:

! ! ! Equation 3.2: � �!! × 1 − � + � �!!× � = � �!!

δ2H values of n-alkanes synthesized during t’ were derived using the leaf water δ2H values during the irrigation with highly 2H-enriched water and applying a mean biosynthetic

fractionation (εbio) value for C3 grasses of –197.8 (Gamarra et al., in review) according to: 78 Chapter 3: Secondary n-alkane synthesis

! ! Equation 3.3: � �!! = ℇ!"# + 1 × � �!"#$ !"#$% + 1 − 1

Finally, we divided the fraction of n-alkanes synthesized between at t’ by the number of days between t1 and t2 and multiplied this value with 100 to determine the percentage of secondary leaf wax synthesis on mature leaves per day.

3.4. Results

3.4.1. Leaf water evaporative 2H-enrichment

2 Leaf water δ H values at t1, before tracer application, varied among the six species in the wet chamber between -37.4‰ and -24.4‰ with a species mean of -29.6‰ and between -8.9‰ and 3.6‰ with a species mean of 0.4‰ in the dry chamber (Figure 3.1 and Table S3.1). The degree of variability of leaf water δ2H values was larger for the dry (SD 11.5‰) than the wet

2 (SD 4.8‰) treatment. At t2, leaf water δ H values increased substantially in both treatments due to irrigation with 2H-enriched water (Table 3.3 and S3.1). Leaf water values in the wet chamber ranged between 260.8‰ and 508.7‰ with a species mean of 378.6‰ and in the dry chamber between 510.0‰ and 735.3‰ and a species mean of 632.0‰ (Figure 3.1 and

Table S3.1). Differences in the leaf water δ2H values were significant across species, between

RH treatments and between sampling times (Table 3.3).

3.4.2. Leaf wax n-alkane concentrations

From all long chained n-alkanes, the most abundant compounds in all species were nC29, nC31 and nC33. This pattern was consistent in both RH treatments and at both sampling times

(Table 3.1). Other n-alkane lengths, such as nC23, nC25, nC27 or nC35 were present only in very low concentrations and were thus not considered for stable isotope analyses.

Concentrations for individual n-alkanes for the individual species are reported in Table 1.

When averaged across species leaf wax n-alkane concentrations at t1 in the wet chamber Chapter 3: Secondary n-alkane synthesis 79

were 40.1 µg/g, 73.3 µg/g and 115.1 µg/g for the nC29, nC31 and nC33, respectively and 100.1

µg/g, 193.0 µg/g and 110.3 µg/g for the nC29, nC31 and nC33, respectively in the dry chamber.

At t2 average concentrations across species were 30.5 µg/g, 54.5 µg/g and 66.1 µg/g in the wet chamber while in the dry chamber they were 40.8 µg/g, 83.1 µg/g and 61.3 µg/g for the nC29, nC31 and nC33, respectively. In general leaf wax n-alkane concentrations were significantly higher at t1 compared to t2 and lower in the wet climate chamber than in the dry climate chamber, which was however not significant (Table 3.3).

800 800 t1 ** ** t2 Wet Dry 600 600 H (‰) H (‰) 2 2 δ δ 400 400

200 200 Leaf water Leaf water n.a. n.a. n.a. 0 0

Mean Mean F .rubra F .rubra A. elatius A. elatius L. perenne H. lanatus L. perenne H. lanatus D. glomerataA. odaratum D. glomerataA. odaratum

2 Figure 3.1. Leaf water δ H (‰) at t1 (grey) and t2 (black) of C3 grasses in the wet (left) and dry (right) chamber. Significant differences between leaf water δ2H values were observed among species, t1 and t2 as well as between the dry and the wet chamber. For details see Table 3.3.

3.4.3. Leaf wax n-alkane δ 2H values

2 At t1, CWA leaf wax n-alkane δ H values in the wet treatment ranged from -198.7‰ to -

246.8‰ among species and had an across species mean of -219.3‰. In the dry treatment,

CWA leaf wax n-alkane δ2H values ranged from -181.7‰ to -230.2‰ among species with an 80 Chapter 3: Secondary n-alkane synthesis

across species mean of -196.1‰. On average across species CWA leaf wax n-alkane δ2H values that were 23.2‰ lower in the wet than in the dry treatment (Figure 3.2 and Table

S3.1).

300 t1 Wett1 Dry 300 t2 t2 200 200 H (‰) H (‰) 2 2 δ

100 100 δ n.a. n.a. n.a. 0 0 −100 −100 -Alkane -Alkane n −200 −200 n

Mean Mean F .rubra F .rubra A. elatius A. elatius L. perenne H. lanatus L. perenne H. lanatus D. glomerataA. odaratum D. glomerataA. odaratum

2 Figure 3.2. Leaf wax n-alkane δ H (‰) at t1 (grey) and t2 (black) of C3 grasses. Significant

2 differences between n-alkane δ H values were observed between t1 and t2. For details see Table 3.3.

2 At t2, CWA leaf wax n-alkane δ H values were increased compared to t1. This effect was significant and independent of species and RH treatment (Table 3.3). At t2 CWA leaf wax n-alkane δ2H values were highly variable and ranged across species from -202.3‰ to -9.2‰ with a mean value of -126.2‰ (SD 77.9‰) in the wet treatment and ranged from -110.4‰ to 196.4‰ with a mean value of 5.4‰ (SD 137.5‰) in the dry treatment. Incorporation of

2 2 the H-signal in the CWA leaf wax n-alkane δ H from t1 to t2 thus, correspond to an across species average enrichment of 93.1‰ and 201.5‰ in the wet and dry treatment, respectively. Chapter 3: Secondary n-alkane synthesis 81

3.4.4. Secondary leaf wax n-alkane synthesis

2 2 Following the H-labeling at t2, all plants showed H-incorporation in their leaf wax n-alkanes when compared to t1. Across all species, secondary n-alkane synthesis ranged from 0.09% to

0.86% per day in the wet and 0.27% to 1.09% per day in the dry treatment (Table 3.2 and

Figure 3.3). Mean secondary leaf wax n-alkane synthesis was 0.3% and 0.58% per day for plants in the wet and dry chambers respectively. These differences were, however, not significant (Table 3.3). H. lanatus and D. glomerata in the wet chamber had the lowest secondary n-alkane synthesis with values below 0.2% per day, followed by A. odoratum (wet and dry chambers) with a combined mean value below 0.3% per day. F. rubra (wet and dry chamber) had the maximum secondary n-alkane synthesis with 0.86% per day in the wet and

1.09% per day in the dry treatment.

Table 3.2. Secondary n-alkane synthesis (% per day) and time to reach complete n-alkane regeneration (days) in C3 grasses under wet and dry RH.

Secondary n-alkane Days to regenerate RH synthesis (%/day) SD (days)

L. perenne Wet n.a. n.a. Dry 0.53 0.03 190 A. elatius Wet 0.48 0.19 210 Dry n.a. n.a. D. glomerata Wet 0.17 0.01 588 Dry 0.44 0.18 227 H. lanatus Wet 0.09 0.02 1034 Dry n.a. n.a. A. odoratum Wet 0.32 0.06 316 Dry 0.27 0.11 370 F. rubra Wet 0.86 0.19 116 Dry 1.09 0.31 91 Species Mean Wet 0.3 SD Wet 0.3 Species Mean Dry 0.58 SD Dry 0.36

82 Chapter 3: Secondary n-alkane synthesis

3.5. Discussion

3.5.1. Leaf water δ 2H values

As expected leaf water δ2H values were 2H-enriched in the dry compared to the wet treatment for all species at t1 (Table 3.4 and S3.1). This is because the ratio of atmospheric vapor pressure to leaf internal vapor pressure is a key variable influencing the evaporative

2H-enriched of leaf water, with lower values triggering a stronger effect on the evaporative

2H-enrichment (Craig and Gordon, 1965a; Dongmann et al., 1974; Barbour, 2007; Kahmen et al., 2008). The application of a 2H-tracer in form of irrigation water of 1000‰ increased leaf water δ2H values in the wet chamber to values that ranged between 260.8‰ and 508.7‰ with a mean of 378.6‰ and in the dry chamber between 510.0‰ and 735.3‰ with a mean of 632.0‰ (Figure 3.1). As such, leaf water was 408.2 per mil more enriched at t2 in the wet and 620.0 per mil in the dry chamber at t2. Interestingly, only a fraction of the 1000‰ that we applied as tracer was observed in the leaves. This can also be explained by basic leaf water isotope theory: Leaf water is in consistent exchange with atmospheric water vapor

(Song et al., 2015). The water vapor δ2H values in the chambers ranged between -110‰ and

-90‰ for the wet and dry chamber, respectively (Gamarra et al., in review). As such equilibrium exchange between 2H-enriched leaf water and the 2H-depleted atmospheric vapor will cause leaf water to be less 2H-enriched as the tracer that we had applied. Also, we observed that the effects of the 2H-enriched tracer on leaf water δ2H values were larger in the dry chamber compared to the wet chamber. On average the difference between these treatments was 253.4‰ (Figure 3.1). This can also be explained by basic isotope theory because the effects of water vapor δ2H values on leaf water δ2H values increase with atmospheric moisture (Craig and Gordon, 1965b; Dongmann et al., 1974; Barbour, 2007;

Kahmen et al., 2008). The patterns we describe here are consistent with the tracer-induced Chapter 3: Secondary n-alkane synthesis 83

2H-enrichment patterns described previously by Kahmen et al. (2011) and Gao et al. (2011;

2012).

Table 3.3. Results of an ANOVA testing the effects of species, RH treatment and time of tracer application on (1) leaf water δ2H values, (2) concentrations, (3) n-alkane δ2H values and (4) secondary n-alkane synthesis.

Df F value p-value (1) Leaf water δ2H values Species 5 5.5 <0.001 RH 1 26.1 <0.001 Time 1 621.3 <0.001 groups (error) 55 (2) n-Alkane concentration Species 5 12.6 <0.001 RH 1 6.4 0.014 Time 1 7.2 0.009 groups (error) 62 (3) n-Alkane δ2H values Species 5 3.4 0.009 RH 1 5.9 0.018 Time 1 65.2 <0.001 groups (error) 62 (4) Secondary n-Alkane synthesis Species 5 15.8 <0.001 RH 1 3.1 0.097 groups (error) 18

3.5.2. Leaf wax n-alkane concentrations

The leaf wax n-alkane concentrations that we present here are in a similar range than previously published data (Zhang et al., 2004; Smith and Freeman, 2006; Rommerskirchen et al., 2006; Vogts et al., 2009; McInerney et al., 2011) n-Alkane concentrations of grasses in the wet chamber were lower than in the dry chamber. These findings are in line with results 84 Chapter 3: Secondary n-alkane synthesis

from other studies (Sachse et al., 2006; Hoffmann et al., 2013; Gamarra et al., in review) that have suggested a possible dependency of leaf wax n-alkane concentrations on humidity.

We also found higher leaf wax n-alkane concentrations at t1 than at t2. This pattern was irrespective of treatment and species. Our data thus suggest that between t1 and t2 leaves of all species had lost n-alkanes from their leaves. Although we cannot provide a mechanistic explanation for such losses, previous studies have shown that leaf wax n-alkane concentrations can vary substantially over time. Gao et al. (2011), for example, found variable concentrations in grasses during a day lapse, in a wide range from 40 to 140 µg/g.

Gao et al. (2011) and also Jetter et al. (2006) have thus suggested that lipid waxes such as n- alkanes experience constant decay and replacement. Along these lines, our data suggest that the grasses we investigated here show higher rates of decay over replacement, at least during our experimental time. Our data differ, however, from results of Kahmen et al. (2011) who found remarkably constant leaf wax n-alkane concentrations from the dicot P. trichocarpa in their different leaf types fluctuating only from 2.7 to 2.9 mg/g. In this study, leaf wax n-alkane concentration values in fully mature leaves were in general also lower than in young emerging leaves.

3.5.3. Leaf wax n-alkane δ 2H values

2 At t1 CWA leaf wax n-alkane δ H values ranged from -246.8‰ to -198‰ and from -230.2‰ to -181.7‰ across the six species with across species averages of -219.3‰ and -196.1‰ for the wet and dry treatments, respectively (Figure 3.2). These n-alkane δ2H values are in the same range reported previously in the literature not only for plants grown in artificial but also under natural conditions (more references Sachse et al., 2012). After the application of the tracer CWA leaf wax n-alkane δ2H values ranged from -202.3‰ to -9.2‰ and from -

110.4‰ to 196.4‰ at t2 for plants in the wet and dry treatment, respectively. When averaged across species leaf wax n-alkane δ2H values were 93.1‰ and 201.5‰ more Chapter 3: Secondary n-alkane synthesis 85

2 positive than at t1 in the wet and dry treatment, respectively. The increase in n-alkane δ H values in mature grass leaves between t1 to t2 suggests that n-alkanes are synthesized after a grass leaf has emerged from the sheath and has reached maturity. The difference in leaf wax

2 n-alkane δ H values between wet and dry treatments at t2 can be attributed to a set of reasons. Given that the RH treatment in the two chambers resulted in more positive bulk leaf water δ2H values on plants in the dry chamber it is not surprising that their newly regenerated n-alkanes reflect similar 2H-enrichement trends. Another explanation for more

2 positive n-alkane δ H values in the dry chamber at t2 is that secondary leaf wax synthesis is enhanced in the presence of external stressors such as desiccation due to low humidity conditions. In fact, we found higher secondary leaf way synthesis rates in the dry chamber as compared to the wet chamber, although this effect was only marginally significant (Table

2 3.3, see below for details). Importantly, leaf wax n-alkane δ H values at t2 were less enriched

2 in both treatments compared to t1 than the tracer induced leaf water H-enrichment at t2.

Lower tracer-induced 2H-enrichement of n-alkanes compared to leaf water is because leaf waxes sampled at t2 are a mix of n-alkanes present already before the tracer application as well as n-alkanes that newly synthesized during the application of the 2H tracer.

3.5.4. Secondary leaf wax n-alkane synthesis

Our data show that our six C3 grass species had secondary n-alkane synthesis rates that varied across species between 0.09% and 0.86% newly synthesized n-alkanes per day and from 0.27% to 1.09% newly synthesized n-alkanes per day for the wet and the dry chamber, respectively. Our data clearly show that mature grass leaves do not cease the biosynthesis of leaf wax n-alkanes and that this pattern is consistent across different species. However, the rate of secondary synthesis varies by almost one order of magnitude among species. In general, however, the secondary n-alkane synthesis rates are relatively low, compared to the primary synthesis of leaf waxes that occurs during only a few days. We calculated that a 86 Chapter 3: Secondary n-alkane synthesis

complete n-alkane regeneration for the six C3 grass species would take almost three years

(1034 days) for H. lanatus (wet treatment), 588 days for D. glomerata (wet treatment) and for A. odoratum (wet and dry treatment), D. glomerata (dry treatment) and A. elatius (wet treatment) 316, 370, 227 and 210 days, respectively. Only F. rubra (dry treatment) had rates of secondary n-alkane synthesis that resulted in complete regeneration values of only 91 days.

1.5 1.5 Wet Dry

1.0 1.0

0.5 0.5

n.a. n.a. n.a.

-Alkane secondary synthesis (%/day) 0 0 -Alkane secondary synthesis (%/day) n n

Mean Mean F .rubra F .rubra A. elatius H. lanatus A. elatius H. lanatus L. perenne L. perenne D. glomerataA. odaratum D. glomerataA. odaratum

Figure 3.3. n-Alkane secondary synthesis rates (% per day) of C3 grasses. Significant differences in secondary leaf wax n-alkane synthesis were observed among species but not between the dry and the wet chamber. For details see Table 3.3.

As such, our study suggests that the secondary synthesis of n-alkanes in grasses exists but is relatively low at least under the stable environmental conditions that our plants grew in. n-Alkane regeneration could be more important especially when plants grow in natural environments, where they are exposed to a multitude of stresses (Baker, 1974; Cameron,

2005; Shepherd and Griffiths, 2006). This is in particular, since our data suggest enhanced secondary n-alkane synthesis in the dry chamber compared to the wet chamber, although Chapter 3: Secondary n-alkane synthesis 87

this effect was only marginally significant (Table 3.2 and 3.3). By comparison, the literature shows mixed results on secondary n-alkane synthesis or n-alkane regeneration. Gao et al.

(2012) modeled a 100% regeneration of n-alkanes in Phleum pratense in only a few weeks.

On the other hand, Gao et al. (2012) and Kahmen et al. (2011) found for F. americana and P. trichocarpa that secondary n-alkane synthesis occurs at insignificant levels (or not at all for

Kahmen et al., 2011).

Our sensitivity analysis shows that uncertainties in the model input variables used in the binary isotope model do not have a large effect on the calculated secondary n-alkane synthesis rates. This shows, for example, that even with uncertainties in leaf water δ2H values of one order of magnitude, estimated rates of secondary n-alkane synthesis would

have been affected by less than 0.88% per day. In C3 grasses described values of εbio ranges from -173‰ to -230‰ (Gamarra et al., in review). An uncertainty of 50‰ in εbio would result in a maximum change of secondary n-alkane synthesis of only 0.11% per day. An increase in leaf wax n-alkane δ2H value of 100‰ would result in a change of 1.71% per day in the secondary n-alkane synthesis.

Since secondary n-alkane synthesis in the dry chamber was higher than in the wet chamber, although this effect was only marginally significant (Table 3.2 and 3.3), more positive δ2H values in leaf wax n-alkanes in the dry chamber compared to the wet chamber are considered in our calculations of secondary n-alkane synthesis, as the calculation of secondary leaf wax synthesis is approached from the leaf water δ2H values in the different climate chambers.

In combination with findings of Gao et al. (2011; 2012) and Kahmen et al. (2011), our study suggests that grasses produce cuticular lipids such as n-alkanes after leaf maturity, although the magnitude of this effect has been reported differently in Gao et al. (2012)

(although only for one species) that we report here. This is in contrast to dicots, where 88 Chapter 3: Secondary n-alkane synthesis

Kahmen et al. (2011) and Gao et al. (2012) found no secondary leaf wax synthesis and suggests fundamental differences between the biosynthesis of leaf wax n-alkane between dicots and grasses. Interestingly, Sachse et al. (2010) also suggest no secondary leaf wax synthesis in the grass Hordeum sativa (barley) after leaf maturation. Sachse et al. (2010) used, however, the natural variability of n-alkane δ2H values in barley leaves over the course of a season. This variability is probably too low to detect secondary leaf wax synthesis rates that range approximately between 0.1 and 1% per day. In turn this highlights that although we found secondary leaf wax synthesis in mature grass leaves, this effect has probably only little implications for the interpretation of grass derived leaf wax n-alkane δ2H values in environmental studies. This is, because the majority of n-alkanes on a mature grass leaf will be primary n-alkanes that are synthesized at the intercalary meristem at the base of the leaf, where the leaf water experiences only reduced evaporative 2H-enrichment (Kahmen et al.,

2013; Gamarra et al., in review). Higher secondary leaf wax n-alkane rates, would strengthen the effect of leaf water evaporative 2H-enrichment in the leaf wax n-alkane δ2H signal because secondary n-alkanes would be synthesized in enriched bulk leaf water at the evaporative sites and would thus record the full extension of such isotope enrichment.

3.6. Conclusion

Our study shows that there is occurrence of leaf wax n-alkane synthesis in grasses after maturity. Such secondary synthesis however varies across species and it is relatively low compared with the primary n-alkane synthesis occurring when the leave emerges at the intercalary meristem. However, regeneration values could be more important in naturally occurring conditions as it is the case of grasslands where seasoning is long and in places that are exposed to environmental stressed conditions such as extreme droughts, abrasive wind or high UV light exposure. Although leaf wax n-alkane δ2H values are mostly recorded early Chapter 3: Secondary n-alkane synthesis 89

in the ontogeny stages of the leaf development, we suggest that in some cases n-alkane δ2H values would also have information of environmental or physiological conditions occurring after leaf maturation. This is especially the case of species that continue regenerating n- alkanes during maturity, as it is the case of F. rubra. Our findings thus have important implications in paleoecological studies. A continuous regeneration of n-alkane in grasses would partially explain the variability found in δ2H values of lipid biomarkers in sedimentary records as they would reflect processes occurring at different moments during the growth season and not mechanisms occurring in shorter periods of early plant development.

3.7. Acknowledgements

The authors would like to acknowledge the Grassland group at the Department of

Environmental System Sciences at ETH Zurich (D-USYS) and to the Department of

Environmental Sciences – Botany, University of Basel, for their contribution as host laboratories. ERC starting grant COSIWAX (ERC-2011-StG Grant Agreement No. 279518) sponsored BG and AK in the development of this study.

90 Chapter 3: Secondary n-alkane synthesis

3.8. Supplementary material

Table S3.1. δ2H values of leaf water and CWA leaf wax n-alkanes of C3 grasses under wet and dry RH at t1 and t2.

2 2 t1 δ H leaf δ H Humidity water SD n-alkane SD (‰) (‰) L. perenne Wet n.a. n.a. Dry -8.9 4.4 -183.7 3.3 A. elatius Wet -24.4 1.1 -226.9 10.8 Dry n.a. n.a. D. glomerata Wet -29.1 1.1 -198.7 4.0 Dry 3.6 2.2 -181.7 4.7 H. lanatus Wet -28.1 1.0 -216.7 7.1 Dry n.a. n.a. A. odoratum Wet -37.4 2.4 -207.5 2.8 Dry -8.5 3.2 -188.7 19.4 F. rubra Wet -28.8 0.3 -246.8 5.3 Dry 15.4 1.2 -230.2 4.7 Species Mean Wet -29.6 -219.3 SD 4.8 18.6 Species Mean Dry 0.4 -196.1 SD 11.5 22.9

2 2 t 2 δ H leaf δ H Humidity water SD n-alkane SD (‰) (‰) L. perenne Wet n.a. n.a. Dry 735.3 23.9 12.5 24.8 A. elatius Wet 508.7 52.1 -89.7 73.8 Dry n.a. n.a. D. glomerata Wet 260.8 75.5 -179.7 14.0 Dry 588.5 15.9 -77.0 7.4 H. lanatus Wet 332.4 69.5 -202.3 2.0 Dry n.a. n.a. A. odoratum Wet 326.5 10.5 -150.3 11.8 Dry 510.0 28.0 -110.4 21.6 F. rubra Wet 464.3 30.6 -9.2 66.5 Dry 694.3 18.9 196.4 132.3 Species Mean Wet 378.6 -126.2 SD 103.7 77.9 Species Mean Dry 632.0 5.4 SD 102.2 137.5

Chapter 3: Secondary n-alkane synthesis 91

3.9. References

Baker EA. 1974. The influence of environment on leaf wax development in Brassica oleracea var. gemmifera. New Phytologist 73: 955–966.

Barbour MM. 2007. Stable oxygen isotope composition of plant tissue: a review. Functional Plant Biology 34: 83.

Cameron KD. 2005. Increased Accumulation of Cuticular Wax and Expression of Lipid Transfer Protein in Response to Periodic Drying Events in Leaves of Tree Tobacco. Plant Physiology 140: 176–183.

Christie WW, Han X. 2010. Lipid Analysis: Isolation, Separation, Identification and Lipidomic Analysis. 4. The Oily Press.

Craig H., Gordon L.I. 1965. Deuterium and oxygen 18 variations in the ocean and the marine atmosphere. Pisa, Italy: Consiglio Nationale Delle Ricerche Laboratorio Di Geologia Nucleare.

Dongmann DG, Nürnberg HW, Förstel H, Wagener K. 1974. On the enrichment of H2 18O in the leaves of transpiring plants. Radiation and Environmental Biophysics 11: 41–52.

Feakins SJ, Sessions AL. 2010. Controls on the D/H ratios of plant leaf waxes in an arid ecosystem. Geochimica et Cosmochimica Acta 74: 2128–2141.

Gao L, Burnier A, Huang Y. 2011. Quantifying instantaneous regeneration rates of plant leaf waxes using stable hydrogen isotope labeling. Rapid Communications in Mass Spectrometry 26: 115–122.

Gao L, Tsai Y-J, Huang Y. 2012. Assessing the rate and timing of leaf wax regeneration in Fraxinus americana using stable hydrogen isotope labeling. Rapid Communications in Mass Spectrometry 26: 2241–2250.

Gehre M, Geilmann H, Richter J, Werner RA, Brand WA. 2004. Continuous flow2H/1H and18O/16O analysis of water samples with dual inlet precision. Rapid Communications in Mass Spectrometry 18: 2650–2660.

Hauke V, Schreiber L. 1998. Ontogenetic and seasonal development of wax composition and cuticular transpiration of ivy (Hedera helix L.) sun and shade leaves. Planta 207: 67– 75.

Hoffmann B, Kahmen A, Cernusak LA, Arndt SK, Sachse D. 2013. Organic Geochemistry. Organic Geochemistry 62: 62–67.

Jetter R, Schäffer S. 2001. Chemical composition of the Prunus laurocerasus leaf surface. Dynamic changes of the epicuticular wax film during leaf development. Plant Physiology 126: 1725–1737.

Jetter R, Kunst L, Samuels AL. 2006. Composition of plant cuticular waxes. In 'Biology of the plant cuticle,.(Eds M Riederer, C Müller) pp. 145-181. 92 Chapter 3: Secondary n-alkane synthesis

Kahmen A, Dawson TE, VIETH A, Sachse D. 2011. Leaf wax n-alkane δD values are determined early in the ontogeny of Populus trichocarpa leaves when grown under controlled environmental conditions. Plant, Cell & Environment 34: 1639–1651.

Kahmen A, Schefuß E, Sachse D. 2013. Leaf water deuterium enrichment shapes leaf wax n- alkane dD values of angiosperm plants I: Experimental evidenceand mechanistic insights. Geochimica et Cosmochimica Acta 111: 39–49.

Kahmen A, Simonin K, Tu KP, Merchant A, Callister A, Siegwolf R, Dawson TE, Arndt SK. 2008. Effects of environmental parameters, leaf physiological properties and leaf water relations on leaf water δ 18O enrichment in different Eucalyptus species. Plant, Cell & Environment 31: 738–751.

McInerney FA, Helliker BR, Freeman KH. 2011. Hydrogen isotope ratios of leaf wax n-alkanes in grasses are insensitive to transpiration. Geochimica et Cosmochimica Acta 75: 541–554.

Newberry SL, Kahmen A, Dennis P, Grant A. 2015. n-Alkane biosynthetic hydrogen isotope fractionation is not constant throughout the growing season in the riparian tree Salix viminalis. Geochimica et Cosmochimica Acta 165:75-85.

Pedentchouk N, Sumner W, Tipple B, Pagani M. 2008. δ 13 C and δD compositions of n- alkanes from modern angiosperms and conifers: an experimental set up in central Washington State, USA. Organic Geochemistry 39:1066-1071.

Peters KE, Walters CC, Moldowan JM. 2005. The Biomarker Guide: Biomarkers and isotopes in petroleum systems and Earth history. Cambridge University Press.

Rommerskirchen F, Plader A, Eglinton G, Chikaraishi Y, Rullkötter J. 2006. Chemotaxonomic significance of distribution and stable carbon isotopic composition of long-chain alkanes and alkan-1-ols in C4 grass waxes. Organic Geochemistry 37: 1303–1332.

Sachse D, Billault I, Bowen GJ, Chikaraishi Y, Dawson TE, Feakins SJ, Freeman KH, Magill CR, McInerney FA, van der Meer MTJ, et al. 2012. Molecular Paleohydrology: Interpreting the Hydrogen-Isotopic Composition of Lipid Biomarkers from Photosynthesizing Organisms. Annual Review of Earth and Planetary Sciences 40: 221–249.

Sachse D, Dawson TE, Kahmen A. 2015. Seasonal variation of leaf wax n-alkane production and δ2H values from the evergreen oak tree, Quercus agrifolia. Isotopes in environmental and health studies 51: 124–142.

Sachse D, Gleixner G, Wilkes H, Kahmen A. 2010. Leaf wax n-alkane δD values of field-grown barley reflect leaf water δD values at the time of leaf formation. Geochimica et Cosmochimica Acta 74: 6741–6750.

Sachse D, Kahmen A, Gleixner G. 2009. Significant seasonal variation in the hydrogen isotopic composition of leaf-wax lipids for two deciduous tree ecosystems (Fagus sylvativa and Acerpseudoplatanus). Organic Geochemistry 40: 732–742. Chapter 3: Secondary n-alkane synthesis 93

Sachse D, Radke J, Gleixner G. 2006. δD values of individual n-alkanes from terrestrial plants along a climatic gradient – Implications for the sedimentary biomarker record. Organic Geochemistry 37: 469–483.

Schefuß E, Schouten S, Schneider RR. 2005. Climatic controls on central African hydrology during the past 20,000 years. Nature 437: 1003–1006.

Shepherd T, Wynne Griffiths D. 2006. The effects of stress on plant cuticular waxes. New Phytologist 171: 469–499.

Smith FA, Freeman KH. 2006. Influence of physiology and climate on δD of leaf wax n- alkanes from C3 and C4 grasses. Geochimica et Cosmochimica Acta 70: 1172–1187.

Song X, Simonin KA, Loucos KE, Barbour MM. 2015. Modelling non-steady-state isotope enrichment of leaf water in a gas-exchange cuvette environment. Plant, Cell & Environment 38: 2618–2628.

Tierney JE, Russell M. Northern Hemisphere Controls on Tropical Southeast African Climate Science 2008, 322.

Tipple BJ, Berke MA, Doman CE, Khachaturyan S, Ehleringer JR. 2013. Leaf-wax n-alkanes record the plant–water environment at leaf flush. Proceedings of the National Academy of Sciences 110: 2659–2664.

Vogts A, Moossen H, Rommerskirchen F, Rullkötter J. 2009. Organic Geochemistry. Organic Geochemistry 40: 1037–1054.

Werner RA, Brand WA. 2001. Referencing strategies and techniques in stable isotope ratio analysis. Rapid Communications in Mass Spectrometry 15: 501–519.

West JB, HilleRisLambers J, Lee TD, Hobbie SE, Reich PB. 2005. Legume species identity and soil nitrogen supply determine symbiotic nitrogen-fixation responses to elevated atmospheric [CO2]. New Phytologist 167: 523–530.

Zhang Y, Togamura Y, Otsuki K. 2004. Study on the n-alkane patterns in some grasses and factors affecting the n-alkane patterns. The Journal of Agricultural Science 142: 469– 475.

94 Chapter 3: Secondary n-alkane synthesis

Chapter 4: Organ δ2H values 95

Chapter 4

Concentrations and δ 2H values of cuticular n-alkanes vary significantly among plant organs, species and habitats in grasses from an alpine and a temperate European grasslands

Bruno Gamarra1 and Ansgar Kahmen1,2

1 Physiological Plant Ecology, Institute of Agricultural Sciences, ETH Zürich 2 Sustainable Land Use, Department of Environmental Sciences – Botany, University of Basel

Chapter 3 was published in the scientific journal Oecologia, March 2015

96 Chapter 4: Organ δ2H values

4.1. Abstract n-Alkanes are long chained hydrocarbons contained in the cuticle of terrestrial plants. Their hydrogen isotope ratios (δ2H) have been used as a proxy for environmental and plant ecophysiological processes. Calibration studies designed to resolve the mechanisms that determine the δ2H values of n-alkanes have exclusively focused on n-alkanes derived from leaves. It is, however, unclear in which quantities n-alkanes are produced also by other plant organs such as roots or inflorescences and if different plant organs produce distinct n-alkane δ2H values. To resolve these open questions, we sampled leaves, sheaths, stems, inflorescences and roots from a total of 15 species of European C3 grasses in a temperate and an alpine grassland in

Switzerland. Our data show slightly increased n-alkane concentrations and n-alkane δ2H values in the alpine compared to the temperate grassland. More importantly, inflorescences had typically much higher n-alkane concentrations than other organs while roots had very low n-alkane concentrations. Most interestingly the δ2H values of the carbon autonomous plant organs leaves, sheaths and stems were in general depleted compared to the overall mean δ2H value of a species while non-carbon autonomous organs such as roots and inflorescences show δ2H values that are higher compared to the overall mean δ2H value of a species. We attribute organ-specific

δ2H values to differences in the H-NADPH biosynthetic origin in different plant organs as a function of their carbon relations. Finally, we employed simple mass balance calculations to show that leaves are in fact the main source of n-alkanes in the sediment. As such, studies assessing the environmental and physiological drivers of n-alkanes that focus on leaves produce relationships that can be employed to interpret the δ2H values of n-alkanes derived from sediments. This is, despite the significant differences that we find among the δ2H values in the different plant organs. Our study brings new insights into the natural variability of n-alkane δ2H values and has implications for the interpretation of n-alkane δ2H values in ecological and paleohydrological research. Chapter 4: Organ δ2H values 97

4.2. Introduction

H, C and O isotope ratios of plant-originated material correlate with environmental variables such as temperature and precipitation (Craig and Gordon, 1965; Dawson and Ehleringer, 1993).

The stable isotope composition of plant materials has thus been used with success to investigate environmental and physiological processes (Dawson et al., 2002). While oxygen and in particular carbon isotope analysis of plant materials are now widely applied in plant ecological research, the hydrogen isotope composition of plants is rarely considered. This is because of methodological limitations and the difficulty to analyze hydrogen isotopes of bulk plant materials with sufficient accuracy. In particular oxygen-bound hydrogen atoms, e.g. in cellulose, can rapidly exchange with hydrogen from ambient moisture, making it difficult to analyze and interpret hydrogen isotopes in plants on a routine basis.

With the introduction of compound-specific stable isotope analyses it is now possible to determine the hydrogen isotope composition (δ2H) of defined compounds, such as leaf wax lipids and in particular leaf wax n-alkanes (Burgoyne and Hayes, 1998; Hilkert et al., 1999). These compounds are hydrocarbons that are synthesized in the cuticle of terrestrial plants (Buschhaus et al., 2007; Eglinton and Eglinton, 2008). In these compounds, all hydrogen atoms are covalently bound to carbon atoms and thus do not exchange with their environment. As such, exchangeable hydrogen atoms do not hinder the interpretation of δ2H signals in these compounds and open the door to new applications of stable hydrogen isotopes in ecological and environmental research. In addition, n-alkanes have low degradation rates and their δ2H values remain unchanged during sedimentary diagenesis. n-Alkanes therefore accumulate in soil and sediments (Eglinton and Eglinton, 2008) and have thus been celebrated as stable and powerful biomarker for contemporary ecological and paleoclimatological research (Sachse et al., 2012).

Sachse et al. (2004) found a linear relationship between δ2H values of meteoric water and plant-derived n-alkanes δ2H values of different plant types across a European north-south transect. Based on this relationship it has been suggested that n-alkane δ2H values could act as 98 Chapter 4: Organ δ2H values

proxy for precipitation δ2H values. However, large scatter is typically observed when plant- derived n-alkane δ2H values are correlated with precipitation δ2H values across environmental gradients (Sachse et al., 2004; Polissar and Freeman, 2010; Sachse et al., 2012; Garcin et al.,

2012). This suggests that in addition to precipitation δ2H values other factors also influence the

δ2H values of n-alkanes. Over the past years there have been increasing research efforts to resolve the drivers of this scatter. In summary, three major drivers of n-alkane δ2H values have been identified: source water δ2H values (Sauer et al., 2001; Liu et al., 2006), leaf water evaporative 2H-enrichment (Smith and Freeman, 2006; Kahmen et al., 2013a and b), and biosynthetic fractionation (Chikaraishi et al., 2004). Based on this conceptual framework, a better interpretation of plant-derived n-alkane δ2H values, e.g. from sediment records, is now possible.

Most research efforts designed to resolve and quantify the mechanisms that determine the amount and δ2H values of n-alkanes have focused on n-alkanes derived from leaves. In contrast, few studies have analyzed if other plant organs produce n-alkanes in amounts that are comparable to those in leaves. Since it has recently been suggested that the majority of sediment or soil n-alkane could derive from roots rather than from leaves (Kuhn et al., 2010;

Gocke et al., 2011; Huang et al., 2011), it remains unclear if sediment derived n-alkanes are truly

“leaf waxes” or rather “whole plant waxes”. What is more, there is insufficient information if the drivers of n-alkane δ2H values that have now been described for leaves also apply to n-alkane

δ2H values derived from other plant organs such as sheaths, stems, inflorescences or roots. We know from previous studies that there can be significant differences in carbon isotope ratios among different plant organs (e.g. Cernusak et al., 2009). Early work by Ziegler et al., (1976) on plant bulk hydrogen isotope composition indicated that similar within plant variability could also be true for δ2H values. As such, studies assessing the mechanistic drivers of n-alkane δ2H values need to be expanded from being solely leaf focused to include other plant organs in order to determine if systematic within plant variability of n-alkane concentrations and δ2H values exists Chapter 4: Organ δ2H values 99

and to improve our interpretation of δ2H values of plant-derived compounds in the environment.

In the work we present here, we investigated the concentration and δ2H values of n- alkanes in different plant organs such as leaves, stems, inflorescences and roots of 15 different temperate and alpine C3 grasses that grew in two contrasting environments in Switzerland. The goal of our study was (i) to assess if n-alkanes are synthesized primarily in the cuticle of leaves or if other plant organs such as roots or inflorescences also produce large amounts of n-alkanes as has been recently suggested, and (ii) to assess if systematic differences exist in the hydrogen isotope composition of n-alkanes derived from different plant organs. With these two objectives, our study aims to contribute to a more robust interpretation of n-alkane δ2H values in plants and sedimentary records.

4.3. Materials and Methods

4.3.1. Sampling

To determine the concentrations and δ2H values in different plant organs, we sampled a total of

15 European C3 grass species that grew either in a temperate or in an alpine grassland in

Switzerland. We selected the two sites based on their differences in environmaral conditions and their partly overlapping species composition: The study sites were in Ennetbaden, Aargau for the temperate grassland and Alp Weissenstein, Graubünden for the apline grassland. The research station Alp Weissenstein is located at 46°34’N, 9°47’E and an elevation of 1978 meters above sea level (m.a.s.l.). The growing season at this location lasts from mid-June to September.

The annual precipitation is 918 mm and the mean annual temperature 2.3°C. The research site at Ennetbaden is located at 47°29’ N, 8°19’ E and at an elevation of 360 m.a.s.l.. The growing season lasts from March to October. The annual precipitation is 1057 mm, and mean annual temperature 9°C. 100 Chapter 4: Organ δ2H values

We sampled the most abundant species at each site. From the alpine grassland these were: Dactylis glomerata, Poa alpina, Phleum raeticum, Festuca rubra, Nardus stricta, Briza media, Sesleria caerulea and Deschampsia cespitosa. From the temperate grassland we sampled

Brachypodium pinnatum, Arrhenatherum elatius, Dactylis glomerata, Festuca rubra, Phleum pratense, Lolium perenne and Lolium multiflorum. For each species about 20 to 30 individuals were collected and bulked in order to obtain sufficient plant material. All individuals of a species that we collected grew within a proximity of 20 meters. The total sampling area was of about 50 x 50 m for Ennetbaden and of about 200 x 200 m for Alp Weissenstein. To assure plants were fully matured,they were collected during peak growing season at both sites (in Ennetbaden the

27-Jun-2012 and in Alp Weissenstein the 22-Jul-2013). After their collection, plant material was dried for 24 h at 60°C in a drying oven. Fertile grass tillers were then divided into their five main organs: leaves, stems, sheaths, roots and inflourescences, while infertile tillars without inflorescences were sectioned into vegetative leaves, vegetative stems and vegetative sheaths.

We estimated that about 10 – 50% of the tillers were fertile, with high variability across species.

4.3.2. Lipid purification

After being separated into different organs, approximately 1 gram of plant dried material was clipped into small parts for lipid extraction and purification. Clipped plant parts were transferred into a 40 ml beaker. Total lipids were extracted (TLE) by addition of 30 ml of a mixture of dichloromethane (DCM) and methanol (9:1) and placing the beaker in an ultrasonic bath for 15 minutes (Peters et al., 2005; Christie and Han, 2010). This extraction was done to separate the soluble lipids, which include molecules such as alkyls (including n-alkanes), fatty acids, alcohols and esters from other insoluble components of the plant material, e.g. cellulose or lignine. The long-chained n-alkanes were further purified from other soluble molecules using liquid chromatograpy (LC). To do so we used 6 ml glass columns that were sterilized in a combustion oven at 500°C for 5 hours. ¾ vol of the columns were packed with silica gel 60 (0.040-0.063 mm

Alfa Aesar, Johnson Matthey Company) 99.5% pure. The packed columns were rinsed with 2 Chapter 4: Organ δ2H values 101

volumes of acetone, 2 volumes DCM and 2 volumes of hexane, then the silica gel columns were temperature activated and dried in a desication oven at 60°C overnight. To obtain the fraction containing n-alkanes (fraction 1), the TLE was passed through the silica gel columns with 12 ml of n-hexane (GC-grade). Thereafter, a mixture of hexane and DCM (1:1) was passed through the remaining column to obtain the fatty acids (fraction 2). The remaining apolar material (fraction

3) was obtained by passing a mixture of DCM and methanol (9:1) on the column. For the results reported here, only fraction 1 was considered.

4.3.3. n-Alkane concentration and isotopic analysis n-Alkanes in fraction 1 were identified and quantified using a gas chromatograph (GC; Agilent

Technologies, 7890A) coupled with a flame ionization detector (FID) at the Geologisches Institut

ETH Zurich. Measurements were done on a 30 m column (Agilent, DB-5) with a diameter of

0.250 mm and film tickness of 0.25 µm. The compound quantification was done by peak area comparison with an internal standard (α-androstane). n-Alkane concentrations for entire plants or specific organs are reported throughout this manuscript as µg/g biomass.

We calculated the average chain length (ACL) for each sample to compare the preference of plant organs to synthesize a length of alkane (short or long) as:

Equation 1: ACL = Σ nxC / Σ C

Where ‘n’ is the carbon chain-length (odd only and considering carbon lenghts from nC25 to nC35) and ‘C’ is the concentration of alkane with ‘n’ carbons.

The hydrogen isotope composition of n-alkanes was measured with an isotope ratio mass spectrometer (IRMS; Delta V Plus, ThermoFisher) coupled to a GC (Trace GC Ultra,

ThermoFisher). Compounds were separated in the GC on a 30 m column (Agilent, DB-5) with a diameter of 0.250 mm and film tickness of 0.25 µm. The injector was in splittless mode at a temperature of 270°C. 1µL of each sample with a concentration of 300 ng/µl of the most abundant n-alkane was injected in triplicates. During sample separation the GC oven temperature was first held at 90°C for 2 minutes, then raised to 150°C at 10°C per minute, it 102 Chapter 4: Organ δ2H values

finally reached 320°C at a rate of 4°C per minute. This temperature was held for 10 minutes.

After separation on the GC column, individual compounds were converted to H2 gas in an aluminum oxide reactor at 1420°C. In each sequence, we ran 10 samples in triplicate injections.

An n-alkane standard mixture (A4, provided by A. Schimmelmann, Indiana University) was run at three different concentrations (100 ng/µl, 200 ng/µl, 400 ng/µl) at three times during each sequence. In total this resulted in 42 injections per sequence. Injecting standards in different concentrations revealed that peak sizes below an area of 20 Vs produced unstable δ2H values. As such, sample compounds with peak sizes smaller that 20 Vs were omitted from the analyses. The linear relationship of known and measured δ2H values from the A4 mixture was used to derive sample δ2H values relative to the VSMOW scale. The precision of the instrument was assesed on the basis of an internal laboratory standard (nC29 alkane from Oak leaves) which was measured in all the sequences, its mean standard deviation for all sequences analyses in this study was 9.4 permil. The H3+ factor was calculated at the beginning of each sequence and had a mean value of 2.5 during the analysis for this study.

4.3.4. Data analysis and Statistics

For each plant organ of a species only a single sample consiting of tissue from >20 bulked tillars was analyzed in our study. The reason for bulking the samples of a species was the labor intensive sample preparation and isotope analysis and the fact that the organ of a single individual would not have yielded sufficient amounts of n-alkanes for isotopic analyses. The main experimental unit of our work was thus the “plant organ” with the individual species that we sampled serving as replicates in the statistical analyses (for details see below). To compare

δ2H values of plant organs across species, we determined the concentration weighed mean δ2H values of nC29 and nC31 for each organ of a species. nC29 and nC31 turned out to be the most abundant compounds across the investigated species and allowed reliable hydrogen isotope ratio measurements. To test for significant differences in n-alkane concentrations and δ2H values Chapter 4: Organ δ2H values 103

across plant organs and between research sites, we performed an analysis of variance (ANOVA) with organs and research sites as factors.

To assess if plant organs systematically deviate in their δ2H values from the plants mean

δ2H values in a site, we estimated for each species at each site a “plant mean δ2H value” by averaging the δ2H values of all organs for a species at a site. We then calculated the deviation of n-alkane δ2H values from this mean for each organ of a species. To detect general across species trends in organ δ2H values, we estimated the mean of these deviations for each organ across species at a site. This standardization procedure was necessary because across species variability in δ2H values can be large and exceed within plant variability. To test for significant differences in organs n-alkane δ2H deviation from the species mean, we performed an ANOVA for each of the two sites with organ as factor followed by a Fisher’s Least Significant Differences (LSD) test.

Testing interactions between species and site or organs were not possible due to lack of replication at the species level. All statistical calculations were done using the statiscal package R version 3.1.1 (http://www.r-project.org).

4.4. Results

4.4.1. n-Alkane composition and concentration

We found that grasses at Alp Weissenstein and Ennetbaden had a strong preference for odd over even long chained n-alkanes, with chains containing 25 to 33 carbon atoms being the most common (Table 4.1, Figures 4.1 and 4.2). Among these, nC29 and nC31 were generally the most abundant n-alkanes at both sites and the sum of nC29 or nC31 among species was always higher than any other nC length except for P. pratense at Ennetbaden where the contribution from nC27 was also important (Table 4.1 and Figure 4.2). The ACL for all green organs (leaves, sheaths, stems and vegetative organs) surpasses 30 across all species whereas inflorescences and roots had a species averaged ACL of 28.8 and 29.6 respectively (Table 4.1).

104 Chapter 4: Organ δ2H values

Table 4.1. n-Alkane concentrations (µg/g) and ACL for different organs from grasses at Alp Weissenstein and Ennetbaden.

Alp Weisenstein C25 C27 C29 C31 C33 C25-C33 ACL D. glomerata

Leaves 16 13 38 56 30 153 29.9 Sheath 13 12 61 48 10 142 29.4 Stem 3 3 9 9 2 25 29.3 Veg. Leaves 7 7 16 30 20 80 30.2 Veg. Sheath 13 4 9 18 13 57 29.5 Veg. Stem 2 1 3 3 1 9 28.8 Inflorescence 115 130 143 98 23 509 28.1 Roots 3 11 9 12 13 48 29.9 Species mean 128 29.4

P. alpina Leaves 20 8 13 44 117 202 31.3 Sheath 6 7 24 72 55 163 31.0 Stem 1 3 17 41 10 72 30.5 Veg. Leaves 23 9 14 36 91 173 30.9 Veg. Sheath 5 10 27 46 25 113 30.3 Veg. Stem 8 13 46 110 31 208 30.4 Inflorescence 93 101 133 151 60 538 28.9 Roots 5 6 12 19 9 51 29.8 Species mean 190 30.4

P. raeticum Leaves 18 9 24 88 48 188 30.5 Sheath 9 9 54 121 23 217 30.3 Stem 2 3 46 92 5 148 30.3 Veg. Leaves 11 8 33 118 47 217 30.7 Veg. Sheath 3 4 21 32 8 68 30.1 Veg. Stem 2 4 32 61 8 107 30.3 Inflorescence 165 208 262 185 29 849 28.3 Roots 1 4 6 9 6 26 30.1 Species mean 228 30.1

F. rubra Leaves 15 13 157 324 118 626 30.6 Sheath 3 11 325 406 103 849 30.4 Stem 1 5 90 84 22 202 30.2 Veg. Leaves 14 15 120 328 128 603 30.8 Veg. Sheath 6 8 42 117 61 233 30.9 Veg. Stem n.a. n.a. n.a. n.a. n.a. n.a. n.a. Inflorescence 127 180 205 199 31 742 28.5 Roots 2 12 13 15 8 50 29.7 Species mean 472 30.2

N. stricta Leaves 5 11 133 204 62 415 30.5 Sheath 8 27 352 787 292 1465 30.8 Stem 13 28 333 815 292 1481 30.8 Veg. Leaves 5 13 200 260 62 541 30.3 Veg. Sheath 4 12 151 400 181 749 31.0 Veg. Stem 4 9 112 293 110 527 30.9 Inflorescence 34 108 406 562 116 1226 30.0 Roots 3 31 31 30 11 107 29.3 Species mean 814 30.5

B. media Leaves 30 10 20 140 77 277 30.6 Sheath 21 25 88 60 16 210 29.2 Stem 12 11 26 13 3 65 28.5 Veg. Leaves 47 18 52 107 41 265 29.6 Veg. Sheath 113 72 251 389 80 905 29.6 Veg. Stem 118 96 257 240 38 749 29.0 Inflorescence 120 102 116 137 6 482 28.2 Roots 7 13 23 43 15 102 29.9 Species mean 382 29.3

S. caerulea Leaves 12 42 835 182 12 1083 29.3 Sheath 10 40 3 488 96 637 31.0 Stem 2 20 703 107 3 835 29.2 Veg. Leaves 6 25 502 309 28 871 29.8 Veg. Sheath 12 18 150 82 10 273 29.4 Veg. Stem 5 18 328 311 38 700 30.0 Inflorescence 21 68 292 147 157 685 30.0 Roots 0 3 18 12 2 34 29.7 Species mean 640 29.8

D. cespitosa Leaves 15 15 134 325 97 587 30.6 Sheath 10 23 321 760 97 1211 30.5 Stem 14 28 395 1125 151 1714 30.6 Veg. Leaves 53 38 266 602 246 1204 30.6 Veg. Sheath 9 13 84 202 57 364 30.6 Veg. Stem 6 9 47 104 21 187 30.3 Inflorescence 236 297 594 570 59 1756 28.9 Roots 6 8 25 48 16 104 30.2 Species mean 891 30.3

Mean values at Alp

Weisenstein Leaves 16 15 169 170 70 441 30.2 Sheath 10 19 153 343 87 612 30.6 Stem 6 13 202 286 61 568 30.3 Veg. Leaves 21 17 150 224 83 494 30.3 Veg. Sheath 21 18 92 161 54 345 30.2 Veg. Stem 21 21 118 160 35 355 29.9 Inflorescence 114 149 269 256 60 848 29.0 Roots 3 11 17 24 10 65 29.8 Site mean 466 30.1

Chapter 4: Organ δ2H values 105

Table 4.1. (Continuation)

Ennetbaden C25 C27 C29 C31 C33 C25-C33 ACL B. pinnatum

Leaves 0 9 10 25 4 48 30.0 Sheath 0 3 17 41 4 66 30.4 Stem 0 1 6 12 0 19 30.2 Veg. Leaves 2 11 15 39 6 73 30.0 Veg. Sheath 0 8 25 52 5 90 30.2 Veg. Stem 0 0 7 13 1 20 30.5 Inflorescence 31 107 60 51 7 256 28.2 Roots 0 1 1 1 0 3 29.3 Species mean 72 29.9

A. elatius Leaves 23 10 26 69 82 210 30.7 Sheath 12 12 14 37 51 126 30.6 Stem 3 4 29 44 4 83 30.0 Veg. Leaves 15 7 17 49 62 150 30.8 Veg. Sheath 15 15 39 71 37 176 30.1 Veg. Stem 3 6 29 54 6 98 30.1 Inflorescence 28 56 114 141 42 380 29.6 Roots 3 3 3 5 4 17 29.4 Species mean 155 30.2

D. glomerata Leaves 19 11 18 34 26 108 29.7 Sheath 11 7 16 23 6 64 29.2 Stem 4 2 7 10 2 25 29.2 Veg. Leaves 13 7 13 28 15 75 29.7 Veg. Sheath 26 13 12 9 5 65 27.5 Veg. Stem 34 9 5 3 3 54 26.5 Inflorescence 217 244 271 339 92 1163 28.7 Roots 3 14 10 5 3 34 28.5 Species mean 199 28.6

F. rubra Leaves 6 9 117 71 16 219 29.7 Sheath 3 8 156 62 4 233 29.5 Stem 1 4 51 15 1 72 29.3 Veg. Leaves 3 5 51 49 7 116 29.9 Veg. Sheath 2 5 39 42 8 95 30.0 Veg. Stem n.a. n.a. n.a. n.a. n.a. n.a. n.a. Inflorescence 54 72 126 102 12 364 28.7 Roots 0 3 3 1 0 7 28.6 Species mean 158 29.4

P. pratense Leaves 5 21 5 2 1 33 27.3 Sheath 3 18 4 1 0 26 27.3 Stem 3 4 3 1 0 11 27.6 Veg. Leaves 7 32 8 2 1 50 27.3 Veg. Sheath 4 17 6 1 0 28 27.4 Veg. Stem 2 5 2 2 0 10 27.8 Inflorescence 55 94 82 74 6 312 28.2 Roots 0 1 21 23 0 44 30.0 Species mean 64 27.9

L. perenne Leaves 3 6 29 53 25 116 30.5 Sheath 0 5 69 129 14 216 30.4 Stem 0 1 13 22 2 38 30.3 Veg. Leaves 0 5 31 56 31 123 30.8 Veg. Sheath 0 5 50 95 20 170 30.5 Veg. Stem 0 6 27 44 12 89 30.4 Inflorescence 57 80 67 113 10 327 28.6 Roots 0 1 2 2 1 6 30.3 Species mean 136 30.3

L. multiforme Leaves 3 10 63 86 16 178 30.1 Sheath 0 6 76 98 10 190 30.2 Stem 0 3 24 43 2 71 30.3 Veg. Leaves 0 5 31 56 31 123 30.8 Veg. Sheath 0 5 50 95 20 170 30.5 Veg. Stem 0 6 27 44 12 89 30.4 Inflorescence 91 132 87 106 4 419 28.0 Roots 0 1 2 2 1 6 30.3 Species mean 156 30.1

Mean values at

Ennetbaden Leaves 8 11 38 49 24 131 30.1 Sheath 4 8 50 56 13 132 30.0 Stem 2 2 19 21 2 46 29.8 Veg. Leaves 6 10 24 40 22 101 30.2 Veg. Sheath 7 9 32 52 13 113 30.0 Veg. Stem 6 5 16 27 6 60 29.7 Inflorescence 76 112 115 132 25 460 28.6 Roots 1 3 6 6 1 17 29.4 Site mean 132 29.7

106 Chapter 4: Organ δ2H values

1000 1000 a) D. glomerata b) P. alpina 800 800 nC33 600 nC31 600 nC29 400 nC27 400 nC25 200 200 0 0 1000 1000 g/g) g/g) c) P. raeticum d) F. rubra μ μ 800 800 600 600 400 400

200 200 0 0 1000 2000 e) B. media f) S. caerulea 800 1600 600 1200 400 800 200 400 Alkanes concentration ( concentration − Alkanes Alkanes concentration ( concentration − Alkanes 0 0 n n 2000 2000 g) N. stricta h) D. cespitosa 1600 1600 1200 1200 800 800 400 400 0 0 roots roots stems stems leaves leaves sheaths sheaths veg. stems veg. veg. stems veg. veg. leaves veg. leaves veg. sheaths veg. sheaths inflorescences inflorescences

Figure 4.1. n-Alkane concentrations (µg/g dry leaf weight) of grass organs at Alp Weissenstein. (Note the change in scale on the y-axes for S. caerulea, N. stricta and D. cespitosa).

We found that total n-alkane concentrations varied significantly among species and organs and were marginally significant between sites (Tables 4.1 and 4.3, Figures 4.1 and 4.2). In general, plants had higher n-alkane concentrations at the alpine site Alp Weissenstein where they ranged from 128 to 891 µg/g (Tab. 1) as compared to the temperate site at Ennetbaden, where total n-alkane concentration ranged from 64 to 199 µg/g. Among organs, there was a significant trend across species and sites that inflorescences had the largest n-alkane Chapter 4: Organ δ2H values 107

concentration whereas roots had the lowest amount of n-alkanes. At Alp Weissenstein, the average inflorescence n-alkane concentration was 848 µg/g, while roots had an average concentration of 65 µg/g. Other green organs varied largely: leaves had concentrations that varied between 153 and 1083 µg/g (Table 4.1, Figure 4.1). At Ennetbaden, the average n-alkane concentration for inflorescence was 460 µg/g, while roots had an across species average concentration of only 17 µg/g and leaves varied between 33 and 219 µg/g (Table 4.1, Figure 4.2).

1000 1000 a) B. pinnatum b) A. elatius 800 800 nC33 600 nC31 600 nC29 400 nC27 400 nC25 200 200 0 0 1000 1000 g/g) c) F. rubra d) P. pratense g/g) μ 800 800 μ 600 600 400 400

200 200 0 0 1000 1000 e) L. perenne f) L. multiflorum 800 800

600 600 400 400 200 200 Alkanes concentration ( − Alkanes concentration 0 0 ( − Alkanes concentration n n 2000 g) D. glomerata roots stems 1600 leaves sheaths veg. stems veg. 1200 veg. leaves veg. sheaths

800 inflorescences

400

0

roots stems leaves sheaths veg. stems veg. veg. leaves veg. sheaths

inflorescences

Figure 4.2. n-Alkane concentrations (µg/g dry leaf weight) of grass organs at Ennetbaden. (Note the change in scale on the y-axes for D. glomerata).

108 Chapter 4: Organ δ2H values

2 Table 4.2. Grass organ n-alkane C29, C31 and weighed C29-C31 δ H values at Alp Weissenstein and Ennetbaden sites

Alp Weisenstein C29 C31 C29-C31 Ennetbaden C29 C31 C29-C31 D. glomerata B. pinnatum

Leaves -248 -244.2 -245.9 Leaves -259.4 -261.9 -261.2 Sheath -243.7 -253.6 -248.1 Sheath -256.5 -270.3 -266.2 Stem -271.8 -274.7 -273.2 Stem -273.0 -283.0 -279.7 Veg. Leaves -241.0 -243.6 -242.7 Veg. Leaves -270.5 -271.1 -270.9 Veg. Sheath n.a. -278.0 -278.0 Veg. Sheath -248.6 -269.4 -262.5 Veg. Stem n.a. n.a. n.a. Veg. Stem -271.8 -287.6 -282.1 Inflorescence -214.5 -223.3 -218.1 Inflorescence -205.3 -181.0 -194.1 Roots -230.1 -220.9 -224.8 Roots -205.0 -216.4 -210.0 Species mean -247.2 Species mean -253.4

P. alpina A. elatius Leaves Leaves -259.4 -225.1 -222.1 Sheath -235.0 -236.0 -235.8 Sheath -256.5 -229.7 -229.0 Stem -245.0 -245.0 -245.0 Stem -273.0 -248.6 -247.3 Veg. Leaves n.a. n.a. n.a. Veg. Leaves -270.5 -224.9 -226.1 Veg. Sheath -234.1 -240.2 -237.9 Veg. Sheath -248.6 -247.1 -244.8 Veg. Stem n.a. -252.9 -252.9 Veg. Stem -271.8 -263.2 -261.8 Inflorescence -216.8 -223.6 -220.4 Inflorescence -205.3 -219.2 -218.9 Roots -218.8 -222.2 -220.9 Roots -205.0 -217.8 -213.1 Species mean -235.5 Species mean -232.9

P. raeticum D. glomerata Leaves -254.2 -251.2 -251.8 Leaves -223.3 -225.5 -224.7 Sheath -258.7 -253.1 -254.8 Sheath -244.1 -250.7 -248.0 Stem -285.7 -283.1 -284.0 Stem -245.8 -256.2 -251.7 Veg. Leaves -264.3 -254.0 -256.2 Veg. Leaves -241.0 -243.6 -242.7 Veg. Sheath -246.3 -250.9 -249.1 Veg. Sheath n.a. n.a. n.a. Veg. Stem -275.4 -274.0 -274.5 Veg. Stem n.a. n.a. n.a. Inflorescence -204.8 -215.7 -209.4 Inflorescence -204.6 -209.7 -207.4 Roots -238.3 -238.6 -238.5 Roots -228.4 -216.6 -224.6 Species mean -252.3 Species mean -233.2

F. rubra F. rubra Leaves -256 -255.8 -255.8 Leaves -247.4 -243.5 -245.9 Sheath -257.8 -261.8 -260.0 Sheath -233.9 -227.1 -232.0 Stem -262.7 -263.3 -263.0 Stem -233.5 -224.8 -231.5 Veg. Leaves n.a. -275.2 -275.2 Veg. Leaves -240.7 -240.7 -240.7 Veg. Sheath -247.2 -260.4 -256.9 Veg. Sheath -235.5 -224.0 -229.5 Veg. Stem n.a. n.a. n.a. Veg. Stem n.a. n.a. n.a. Inflorescence -203.5 -201.3 -202.4 Inflorescence -186.2 -186.9 -186.5 Roots -233.7 -222.2 -227.6 Roots -218.4 -201.2 -214.5 Species mean -248.7 Species mean -225.8

N. stricta P. pratense Leaves -267 -267 -267.2 Leaves -183.4 -169.5 -180.0 Sheath -277.6 -268.0 -270.9 Sheath -182.2 -167.1 -178.5 Stem -257.5 -265.6 -263.2 Stem -194.2 -203.6 -196.6 Veg. Leaves -239.1 -247.1 -243.6 Veg. Leaves -201.8 n.a. -201.8 Veg. Sheath -244.9 -254.4 -251.8 Veg. Sheath -222.5 n.a. -222.5 Veg. Stem -251.7 -261.9 -259.1 Veg. Stem n.a. n.a. n.a. Inflorescence -240.1 -256.9 -249.9 Inflorescence -192.3 -189.5 -191.0 Roots -213.0 -217.4 -215.2 Roots -220.2 -220.2 Species mean -252.6 Species mean -198.6

B. media L. perenne Leaves n.a. -248 -248.2 Leaves -221.9 -222.3 -222.2 Sheath n.a. -212.0 -212.0 Sheath -269.6 -267.5 -268.2 Stem -220.7 n.a. -220.7 Stem -253.2 -251.0 -251.8 Veg. Leaves -244.0 -242.7 -243.1 Veg. Leaves -241.6 -233.8 -236.6 Veg. Sheath n.a. -270.4 -270.4 Veg. Sheath -242.9 -245.5 -244.6 Veg. Stem -223.0 -235.7 -229.1 Veg. Stem -253.8 -258.2 -256.5 Inflorescence -205.1 -203.3 -204.1 Inflorescence -211.2 -206.5 -208.3 Roots -224.4 -221.0 -222.2 Roots n.a. -222.9 -222.9 Species mean -231.2 Species mean -238.9

S. caerulea L. multiforme Leaves -252.3 -230 -248.3 Leaves -248.1 -239.7 -243.2 Sheath -272.0 n.a. -272.0 Sheath -257.2 -261.3 -259.5 Stem -279.7 n.a. -279.7 Stem -250.2 -259.8 -256.3 Veg. Leaves -261.9 -241.4 -254.1 Veg. Leaves -241.6 -233.8 -236.6 Veg. Sheath -242.9 -223.6 -236.1 Veg. Sheath -242.9 -245.5 -244.6 Veg. Stem -254.3 -231.2 -243.1 Veg. Stem -253.8 -258.2 -256.5 Inflorescence -200.4 -220.8 -207.2 Inflorescence -219.6 -220.7 -220.2 Roots -236.4 -211.7 -226.6 Roots n.a. -222.9 -222.9 Species mean -245.9 Species mean -242.5

D. cespitosa Leaves -262.1 -256.1 -257.8

Sheath -297.6 -292.6 -294.1

Stem -285.6 -281.2 -282.3

Veg. Leaves -296.8 -282.9 -287.1

Veg. Sheath -288.8 -296.0 -293.9

Veg. Stem -273.5 -277.1 -276.0

Inflorescence -192.6 -190.7 -191.7

Roots -239.7 -252.8 -248.3

Species mean -266.4

Alp Weisenstein Ennetbaden Leaves -256.7 -250.4 -253.6 Leaves -228.3 -226.8 -228.5 Sheath -263.2 -253.9 -256.0 Sheath -238.6 -239.1 -240.2 Stem -263.6 -268.8 -263.9 Stem -242.2 -246.7 -245.0 Veg. Leaves -257.8 -255.2 -257.4 Veg. Leaves -238.2 -241.3 -236.5 Veg. Sheath -250.7 -259.2 -259.3 Veg. Sheath -238.8 -246.3 -241.4 Veg. Stem -255.6 -255.5 -255.8 Veg. Stem -259.6 -266.8 -264.3 Inflorescence -209.7 -217.0 -212.9 Inflorescence -205.4 -201.9 -203.8 Roots -229.3 -225.8 -228.0 Roots -215.6 -216.3 -218.3 Site mean -248.3 Site mean -234.7

Chapter 4: Organ δ2H values 109

4.4.2. n-Alkane δ2H values

2 Concentration-weighted (nC29-nC31) plant-averaged (across organs of a species) n-alkane δ H values ranged from -266‰ to -231‰ at Alp Weissenstein and from -253‰ to -199‰ at

Ennetbaden for the investigated species (Table 4.2, Figures 4.3 and 4.4). In general n-alkane δ2H values were more depleted in plants growing at Alp Weissenstein than at Ennetbaden (Table

4.4). We found significant across organ variability in n-alkane δ2H values which was up to 102‰

(e.g. D. cespitosa) within a species (Figure 4.3, Tables 4.2 and 4.4). In general, inflorescences and roots carried more positive δ2H values than other organs: At Alp Weissenstein, mean δ2H n- alkane values of inflorescences and roots were -213 and -228‰ compared to -254‰ in leaves.

At Ennetbaden, inflorescences and roots had δ2H values of -204 and -218‰ compared to -229‰ in leaves. As such, we observed a general trend for green and carbon autonomous plant organs such as leaves to have more negative δ2H values than organs of the same species that are not carbon autonomous such as inflorescences and roots. This pattern was consistent for both sites and became even more evident, when we calculated the deviation of an individual organ δ2H value from a species mean δ2H value (Table 4.4, Figure 4.5). Across all species in a site, the δ2H values of green organs such as leaves, sheaths and stems, were in general depleted in 2H compared to the plants’ mean δ2H values (Figure 4.5). This was in contrast to inflorescences and roots, which showed always significantly 2H enriched δ2H values. When averaged for each site, we found that δ2H values of leaves, sheaths, stems, vegetative green organs deviated from the species mean δ2H values between -4‰ and -16‰ at Alp Weissenstein and between 3‰ and -

20‰ at Ennetbaden, while inflorescences and roots δ2H values deviated from mean species δ2H values by 35‰ and 19‰ at Alp Weissenstein and 28‰ and 14‰ at Ennetbaden (Figure 4.5).

110 Chapter 4: Organ δ2H values

−150 n.a. n.a. n.a. −150

−200 −200

−250 −250

−300 −300 a) D. glomerata b) P. alpina −350 −350 −150 n.a. −150

−200 −200 H (‰)

−250 −250 H (‰) 2

2 δ −300 −300 δ c) P. raeticum d) F. rubra −350 −350 −150 −150 −200 −200 weighted weighted 31 −250 −250 31

− C −300 −300 − C 29 e) B. media f) S. caerulea 29 C −350 −350 C −150 −150

−200 −200

−250 −250

−300 −300 g) N. stricta h) D. cespitosa −350 −350 roots roots stems stems leaves leaves sheaths sheaths veg. stems veg. veg. stems veg. veg. leaves veg. leaves Plant mean Plant mean veg. sheaths veg. sheaths inflorescences inflorescences

2 Figure 4.3. Organ nC29 - nC31 weighted n-alkane δ H values at Alp Weissenstein. Samples with insufficient n-alkane concentrations for hydrogen isotope analyses are marked (n.a.). Note that for each plant organ only a single sample consiting of tissue from >20 bulked individuals was analyzed.

Chapter 4: Organ δ2H values 111

−150 −150

−200 −200

−250 −250

−300 −300 a) B. pinnatum b) A. elatius −350 −350 −150 n.a. n.a. −150

−200 −200

H (‰) −250 −250 H (‰) 2 2 δ δ −300 −300 c) F. rubra d) P. pratense −350 −350 −150 −150

−200 −200 weighted weighted

31 −250 −250 31

− C −300 −300 − C

29 29 e) L. perenne f) L. multiflorum C −350 −350 C −150 n.a. n.a. roots stems leaves

−200 sheaths veg. stems veg. veg. leaves Plant mean −250 sheaths veg. inflorescences −300 g) D. glomerata −350 roots stems leaves

sheaths veg. stems veg. veg. leaves Plant mean veg. sheaths veg.

inflorescences

2 Figure 4.4. Organ nC29 - nC31 weighted-alkane δ H values at Ennetbaden. Samples with insufficient n-alkane concentrations for hydrogen isotope analyses are marked (n.a.). Note that for each plant organ only a single sample consiting of tissue from >20 bulked individuals was analyzed.

4.5. Discussion

4.5.1. n-Alkane composition and concentration

We found that carbon autonomous organs (e.g. leaves, sheaths, stems and vegetative tissues) had a higher ACL (>30) than non-autonomous organs (e.g. inflorescences and roots) and that inflorescences had on average the lowest ACL values in a plant (Table 1, Figures 1 and 2). We also found a high across species variability in ACL with values ranging from 29.3 to 30.5 at Alp 112 Chapter 4: Organ δ2H values

Weissenstein and from 27.9 to 30.3 at Ennetbaden. ACL values in Alp Weissenstein were in general higher than at Ennetbaden which could be associated with different environmetal conditions at different altitudes. Despite several studies adressed the question, the biological function of ACL is not clear and contradicting results and explanations are found in the literature. For example Sachse et al. (2006) suggest that longer chain lengths indicates better drought resistance while Hoffmann et al. (2013) show variable values. Dodd and Poveda (2003) have suggested that summer drought and winter physiological drought could explain greater chain length at low and high altitudes in plant populations from the Pyrennes. In general, our study supports previous findings that indicate an environmental effect on ACL but the biological functions of a changing ACL remains yet unclear.

We found that average n-alkane concentrations of the individual species (averaged across all organs) ranged from 130 to 890 µg/g at Alp Weissenstein and from 65 to 200 µg/g at

Ennetbaden (Figures 4.1 and 4.2). These values are in the same range as has been reported for other grassland species (Zhang et al., 2004; Rommerskirchen et al., 2006; Vogts et al., 2009). We observed that concentrations were in general higher at the alpine site Alp Weissenstein compared to the temperate site Ennetbaden. These general differences in concentrations between the two sites were consistent among the three species that were sampled at both locations (D. glomerata, F. rubra and P. raeticum). Our data thus suggest that environmental differences between the two sites rather than the identity of species are causing differences in n-alkane concentrations between the two sites. Several previous studies have reported changes in n-alkane concentrations along latitudinal environmental gradients (e.g. Hoffmann et al.,

2013), where n-alkanes tend to increase with increasing aridity. Reports of altitude effects on n- alkane concentrations in plants are, however, rare. Salasoo (1989) found a relative n-alkane increase with altitude in the leaf cuticlar wax composition of Ericaceae. However no absolute amounts were reported in this study. In addition, several previous studies have assessed effects of altitude on cuticle thickness, yet with mixed results. While some studies report increasing Chapter 4: Organ δ2H values 113

cuticle thickness with altitude as response to increased UV-B radiation (Anfodillo et al., 2002), others report reduced cuticle thickness (DeLucia and Berlyn, 1984; Day et al., 1992). It is, however, unclear if cuticle thickness and abundance of n-alkanes per gram dry matter correlate in plants. It would thus be interesting to assess in a future study, if the trend of increased n- alkane abundance with increasing altitude that we report for grasses, can also be found for other taxa in different biomes.

Our study revealed that in grasses n-alkanes are not only present in the cuticle of leaf blades but that other parts of the leaf (sheaths) and other plant organs also contain substantial amounts of n-alkanes. In particular inflorescences had a tendency to have the highest n-alkane concentrations in most species at both sites, in some occasions exceeding the concentrations of leaves by one order of magnitude (Table 4.1, Figures 4.1 and 4.2). Interestingly, the presence of prevalent nC29 and nC31 n-alkanes in inflorescences is accompained by overproportionally high concentrations of shorter length alkanes, such as nC25 or nC27, hence a lower ACL (Table 4.1).

Among others, plants are synthesizing n-alkanes and other wax compounds in order to prevent water losses (Jetter et al., 2006; Buschhaus et al., 2007). Inflorescences are the most critical organ in a plant to ensure reproductive success. Since grass inflorescences are wind pollinated, they typically exceed the grassland canopy and are exposed to the atmosphere above the boundary layer, where evaporative demand is typically higher than within the canopy (Dietrich and Körner, 2014; Jones, 2014). As such, inflorescenses of grasses and possibly also those of other taxonomic groups are prone to dessication (Dietrich and Körner, 2014). Other than leaves, which are redundant and can be replaced in the course of a season, loss of inflorescenses by dessication would result in loss of sexual offspring thus a reduced evolutionary fitness. As such, a strong cuticle and high n-alkane concentrations might be a strategy of the plant to reduce water permeability and make sure that influrescences are dessication resistant, despite being exposed to winds above the canopy boundary layer. In line with our finding, Rommerskirchen et al., 114 Chapter 4: Organ δ2H values

(2006) found that the total n-alkane content in flower heads of Sporobolus sp and Brachiaria sp., was also 3 and 6 times larger than in leaves and stems.

Opposite to inflourescences, we found that roots had only low amounts of n-alkanes

(Table 1, Figures 4.1 and 4.2). In fact, we found that roots were in general the organs with the lowest n-alkane concentrations in all species at both sites. This can be explained because the root tip is meant to be permeable to water to facilitate water uptake by the plant. Here, a water impermeable waxy layer containing high concentrations of n-alkanes would be counter productive. Also, older roots do not have a cuticle to prevent water loss. Instead, older roots develop a suberized bark to prevent water loss as they mature (Fahn, 1990). Given these anatomical and functional differences between root surfaces and other plant organs, the low n- alkane concentrations we found for roots in our study are thus not surprising. Our results confirm several previous studies that have also shown very low n-alkane concentrations in plant roots when compared to other plant organs. Espelie et al. (1980) studied for example the composition of cuticular (from aerial organs) and suberin (from roots) related waxes of seven plant species. In this study, it was found that the amount of total lipid waxes for roots was very small, and varied from only 10 to 50 mg/kg. More recently, Dawson et al. (2000) observed that in 5 grass species (L. perenne, P. trivialis, A. capillaris, F. rubra and F. ovina) the average concentration of nC27-nC33 alkanes in shoots and roots was 187 and 11 mg/kg respectively. It has also been shown that the nC29 and nC31 alkane concentration from aerial organs was usually higher than those of the underground component (Dawson et al., 2000). Li et al. (2007) investigated the wax biosynthesis and enzymatic relations in wild Arabidopsis thaliana. It was found that in stems, leaves, siliques, seed and roots, had wax loads of ca. 860, 80, 1500, 170 and

360 µg/g respectively.

Despite our and previous findings several authors have recently suggested that roots could contribute significant amounts of n-alkanes to soils and sediments (Huang et al., 2011;

Gocke et al., 2011). If true, this would have important implications for the interpretations of Chapter 4: Organ δ2H values 115

isotope ratios of n-alkanes derived from soils or sediments when these are employed in a paleo- environmental context (Sachse et al., 2012). Kuhn et al., (2010) for example have suggested that leaves and roots from C4 grasses are likely to contribute short length alkanes to the n-alkane pool found in soils. In their study, however, the amount of long-chain n-alkanes (C29-C31) in leaves also surpasses those of the roots. Huang et al., (2011) have also suggested that below- ground plant organs could be an important source of lipids in soils and sediments especially for sterols and ketones. Yet in this study, only one species out of ten had higher n-alkane concentrations in roots than in leaves. Gocke et al., (2011) indicated that roots from different plant species would possibly be a source of n-alkanes to soils. However their direct analysis of lipid concentrations also revealed a higher presence of n-alkanes in above-ground tissues.

Although these studies have suggested the possibility of root lipid transfer (especially short length alkanes and other non-alkyl lipids) to soils and sediments, theirs and our results clearly indicate that roots are by far the plant organ with the lowest amount of long chained-alkanes when compared to above ground organs such as leaves and inflorescences.

4.5.2. n-Alkane δ 2H values

We found high across species variability of n-alkane δ2H values (Tables 4.2 and 4.4). It is well established that 3 factors usually determine the n-alkane δ2H values: Source water δ2H values, leaf water 2H-enrichment and biosynthetic 2H-fractionation (Sachse et al., 2012). As expected, n- alkane δ2H values at Alp Weissenstein were generally lower than n-alkane δ2H values at

Ennetbaden. This pattern can mainly be attribuited to the source water in high altitudes which has usually more negative δ2H values than that of lowerlands (Craig, 1961; Siegenthaler and

Oeschger, 1980). 116 Chapter 4: Organ δ2H values

50 50 a) Alp Weissenstein b) Ennetbaden 40 40

30 30

20 20

10 10 ccccc c c 0 bc bc bc bc 0 a ab bc a ab

−10 −10

−20 −20 Organ main deviation (‰) Organ main deviation (‰) Organ main deviation

−30 −30

−40 −40 roots roots stems stems leaves leaves sheaths sheaths veg. stems veg. veg. stems veg. veg. leaves veg. leaves veg. sheaths veg. veg. sheaths veg. inflorescences inflorescences .

Figure 4.5. Deviation of organ δ2H values from the plant’s mean δ2H values averaged seperately for Alp Weissenstein (a) and Ennetbaden (b). δ2H values of individual plant organs were standardized for each species by calculating the deviation of an organ’s δ2H value from the respective species mean δ2H value across all organs. These numbers were then averaged across all species at a site. Letters indicate significant differences between organs. Error bars show one standard deviation from the plant’s mean at a location.

We also found substantial variability in n-alkane δ2H values across different organs within a species (Figures 4.3 and 4.4). When n-alkane δ2H values are standardized across species by calculating the deviation of an organ δ2H value of a respective species mean, it becomes evident that (except for leaves at the Ennetbaden site) all carbon autonomous organs had more negative Chapter 4: Organ δ2H values 117

δ2H values than the plant mean (however if P. pratense was ommited from our analysis, leaves at Ennetbaden would turn from positive to negative values, in Figure 4.5). In contrast, non- carbon autonomous plant organs such as inflorescences and roots always had more positive n- alkane δ2H values than the plant mean at both sites (Figure 4.5). The finding that n-alkanes derived from non-carbon autonomous plant organs are 2H enriched compared to other plant organs is in line with bulk δ13C analysis of different plant organs, where heterotrophic organs are also often more enriched than autotrophic organs in 13C (Cernusak et al., 2009). Although the drivers of these differences are most likely different for carbon and hydrogen isotopes, it is plausible that in both cases biochemical fractionations are responsible for this effect.

There are two possible drivers that determine differences in the n-alkane δ2H values among different plant organs: (i) Leaves and inflorescences might be exposed to microclimates that have an effect on the 2H-enrichment of leaf water. Given that inflorescences have fewer stomata (Larcher, 2001) and at the same time larger loads of n-alkanes than other organs

(preventing cuticular transpiration), they are thus unlikely to experience a larger evaporative 2H enrichment compared to other plant organs, in particular when compared to leaves. In addition, roots also show enriched n-alkane δ2H values compared to other plant organs such as leaves, sheeths and stems. Being located in the soil roots are, however, not exposed to an environment with high evaporative demand. As such, it is unilkely that evaporative enrichment of root water causes the enriched δ2H values of n-alkane that we observed in roots of the plants we investigated here. Alternatively, (ii) differences in biosynthetic fractionation might cause the observed differences in n-alkane δ2H values among different plant organs. In general, n-alkane hydrogen has three sources: water, Acetyl-CoA, and NADPH. Further, NADPH can have two distinct sources of hydrogen. It can come either directly from the light reaction of photosynthesis or from the pentose phosphate pathway where stored carbohydrates are oxidized (Shin, 2004; White et al., 2012). NADPH from the light reaction is more depleted in 2H compared to the NADPH from the pentose phosphate pathway of stored carbohydrates (Luo et 118 Chapter 4: Organ δ2H values

al., 1991; Schmidt et al., 2003). A likely explanation for our results is therefore that n-alkanes formed in leaves, stems and sheaths, i.e. in carbon autonomous plant organs, obtained larger amounts of H derived from NADPH from the light reaction and thus show more negative n- alkane δ2H values. In contrast, the non-green and non-carbon autonumous plant organs roots and inflorescences are decoupled in their tissue formation from photosynthesis and obtain more

2H enriched H that comes from NADPH originating in the pentose phosphate pathway. Zhang et al. (2009) have demonstrated that the source of NADPH is an important source of lipid δ2H values variability in autotrophically and heterotropically grown bacteria. In their study, lipids derived from autotrophically growing bacteria were significantly 2H depleted compared to bacteria that had a heterotropic metabolism. This is directly in line with the patterns we report here, where the green and carbon autonomous organs leaves, sheaths and stems are 2H depleted compared to the non-carbon autonomous plant organs roots and inflorescences. We therefore conclude that different biochemical hydrogen isotope fractionations as a result of organ specific differences in carbon autonomy and carbon allocation cause the organ specific differences in n-alkane δ2H values within a plant.

Interestingly, our findings shed new light on very early analyses of within plant variability of hydrogen isotopes conducted by Ziegler et al. (1976). Just as we report for n-alkanes here,

Ziegler et al. (1976) detected that bulk samples from C autonomous plant organs (such as shoots) have more negative δ2H values than non-autonomous carbon organs (such as roots).

With the new data we report here and in combination with recent insight into the biochemical hydrogen fractionation processes obtained from bacterial growth cultures (Zhang et al., 2009) it is now possible to suggest the biochemical mechanisms that determine this plant internal variability in δ2H values of carbon autonomous and non-carbon autonomous plant organs reported nearly 40 years ago.

Chapter 4: Organ δ2H values 119

4.5.3. Consequences for the interpretation of sediment records

To fully evaluate how n-alkanes derived from different plant organs determine the δ2H values of the sediment record, mass balance assessments of n-alkane production in the individual plant organs are needed. Such assessments are, however, complex because biomass allocation to different plant organs are highly variable across and within species and often controlled by environmental conditions (e.g. Körner, 1991). As such, there is no single ratio that can be applied to biomass allocation to different grass organs and calculating the transfer of organ-specific n- alkanes to soils and sediments is thus difficult. We performed, however rough estimates of biomass allocation to grass organs based on previously published data in the literature.

Depending on the species and environmental variables such as soil moisture, temperature, wind and soil nutrients, the biomass ratio of different organs ranges between 20 – 60% for shoots

(including stems and leaves), 20 to 50% for roots and inflorescences typically contributed less than 20% (Potvin, 1986; Rice et al., 1992; Retuerto and Woodward, 1992; Kalapos et al., 1996;

Rickey and Anderson, 2004; Schwilling et al., 2005).

In addition to records of standing biomass, information on biomass turnover of individual organs, i.e. the amount of standing biomass turning into necromass and litter during a species lifetime, is needed to understand plant organ specific contributions of n-alkanes to the sediment record. Such values are, however, even more diffficult to obtain than standing biomass, in particular for roots. In fact root biomass turnover is a key, yet unanswered question in the global carbon cycle. Data reported in the literature suggest higher biomass and turnovers for leaves than for roots of Arrhenatherum elatius, Dactylis glomerata and Bromus erectus especially under high nutrient conditions (Schläpfer and Ryser, 1996). In addition full root turnover in grasses has been estimated to take 4 to 13 years (Shaver and Billings, 1975; Bell and Bliss, 1977). Thus, turnover rates for individual grass organs seem to be slower for roots than for grasses but are again highly variable and depend on a number of biological and environmental variables.

120 Chapter 4: Organ δ2H values

Table 4.5. Estimated contribution of n-alkanes from the plant organs shoots (leaves and stems), roots and inflorescences to the sediment record. Values for organ-specific biomass allocation are average values estimated from the literature (see text). n-Alkane concentrations are across site averaged values derived from this study.

g) µ

alkane (

g/g) - n µ

Alkane Alkane contribution to the - - Organ derived Organ biomass in reproducing grasses (%) Organ biomass in vegetative grasses (%) Average organ biomass (ecosystem level) (%) n concentration ( per gram of grass n sedminet record (%)

Shoots 50 55 54 286 154.44 83

Roots 40 45 44 41 18.04 10

Inflorescences 10 0 2 654 13.08 7

We used data on biomass allocation to different organs in grasses that are available from the literature and multiplied these with the organ specific n-alkane concentrations that we report here to roughly estimate which plant organs contribute most to the sediment record

(Table 4.5). In summary, these simplistic estimates show that shoots (leaves and stems), inflorescences and roots contribute on average 83, 7 and 10% of the n-alkanes to the sediment record, highlighting that the majority of n-alkanes in the sediment record are in fact derived from shoots. Please note, however, that these are very simplistic “back-of-an-envelope” calculations and that relative contributions will vary in nature, depending on species and ecosystem type as well as environmntal conditions.

Chapter 4: Organ δ2H values 121

4.6. Conclusions

Our study brings new insights into the causes of natural variability of n-alkane δ2H values in plants and has implications for the interpretation of n-alkanes δ2H values in ecological, environmental and paleohydrological research. We show that n-alkane concentrations and 2H isotopic composition differ largely not only across species and sites but also within plant organs.

In contrast to above ground organs such as leaves, shoots and inflorescences, roots have only low n-alkane concentrations. In addition, we found that δ2H values are significantly more negative for carbon autonomous organs such as leaves, sheaths, stems and vegetative organs while non-carbon autonomous inflorescences and roots have more positve δ2H values. We attribute this variability to the carbon metabolism of different plant organs and different associated NADPH sources.

Studies assessing the drivers of δ2H variability in n-alkanes derived from plants have to date been largely conducted on leaves (Sessions et al., 1999, Sachse et al., 2012). n-Alkanes observed in the soil and sediment matrix are, however, a mixture of leaf and inflourescence derived lipids. The real contribution of n-alkanes from different plant organs to the sediment record is difficult to judge. Based on a simple mass balance calculation we conclude, however, that leaves are in fact the main source of n-alkanes in the sediment. As such studies assessing the environmental and physiological drivers of n-alkanes that focus on leaves produce relationships that can be employed to interpret the δ2H values of n-alkanes derived from sediments. This is, despite the significant differences that we find among the δ2H values in the different plant organs.

4.7. Acknowledgements

BG and AK were both funded by the ERC starting grant COSIWAX (ERC-2011-StG - Grant

Agreement N° 279518) to AK. We thank Nadine Brinkmann and Marc-André Cormier for their 122 Chapter 4: Organ δ2H values

help during the fieldwork. We acknowledge Francesca McInerney and one anonymous referee for improving the quality of the manuscript. We thank the Grassland group and the Physiological

Plant Ecology group at the Department of Environmental System Sciences at ETH Zurich (D-USYS) for their contribution as host laboratory and for technical support.

Chapter 4: Organ δ2H values 123

4.8. References

Anfodillo T, Pasqua di Bisceglie D, Urso T. 2002. Minimum cuticular conductance and cuticle features of Picea abies and Pinus cembra needles along an altitudinal gradient in the Dolomites (NE Italian Alps). Tree Physiology 22: 479–487.

Bell KL, Bliss LC. 1978. Root growth in a polar semidesert environment. Can. J. Bot. 56: 2470- 2490

Burgoyne TW, Hayes JM. 1998. Quantitative Production of H2 by Pyrolysis of Gas Chromatographic Effluents. Anal Chem 70: 5136–5141.

Buschhaus C, Herz H, Jetter R. 2007. Chemical Composition of the Epicuticular and Intracuticular Wax Layers on Adaxial Sides of Rosa canina Leaves. Annals of Botany 100: 1557–1564.

Cernusak LA, Tcherkez G, Keitel C. 2009. Why are non-photosynthetic tissues generally 13C enriched compared with leaves in C3 plants? Review and synthesis of current hypotheses. Functional Plant Biol 36:199.

Chikaraishi Y, Naraoka H, Poulson SR. 2004. Carbon and hydrogen isotopic fractionation during lipid biosynthesis in a higher plant (Cryptomeria japonica). Phytochemistry 65: 323–330.

Christie WW, Han X. 2010. Lipid Analysis - Isolation, Separation, Identification and Lipidomic Analysis, 4th edn. Oily Press, Bridgwater, UK

Craig H. 1961. Isotopic variations in meteoric waters. Science 133: 1702–1703.

Craig H, Gordon LI. 1965. Isotopic oceanography: Deuterium and oxygen 18 variations in the ocean and the marine atmosphere, Stable Isotopes in Oceanographic Studies and Paleotemperatures E. In Proceedings of the Spoleto Conference E. Tongiorgi (pp. 9-130)

Dawson LA, Mayes RW, Elston DA, Smart TS. 2000. Root hydrocarbons as potential markers for determining species composition. Plant Cell Environ 23: 743–750.

Dawson TE, Ehleringer JR. 1993. Gender-specific physiology, carbon isotope discrimination, and habitat distribution in boxelder, Acer negundo. Ecology 798–815.

Dawson TE, Mambelli S, Plamboeck AH. 2002. Stable isotopes in plant ecology. Annu Rev Ecol Syst 33: 507–559. 124 Chapter 4: Organ δ2H values

Day TA, Vogelmann TC, DeLucia EH. 1992. Are some plant life forms more effective than others in screening out ultraviolet-B radiation? Oecologia 92: 513–519.

DeLucia EH, Berlyn GP. 1984. The effect of increasing elevation on leaf cuticle thickness and cuticular transpiration in balsam fir. Can J Bot 62: 2423–2431.

Dietrich L, Körner C. 2014. Thermal imaging reveals massive heat accumulation in flowers across a broad spectrum of alpine taxa. Alp Botany 124: 27–35.

Dodd RS, Poveda MM. 2003. Environmental gradients and population divergence contribute to variation in cuticular wax composition in Juniperus communis. Biochemical Systematics and Ecology 31: 1257–1270.

Eglinton TI, Eglinton G. 2008. Molecular proxies for paleoclimatology. Earth and Planetary Science Letters 275: 1–16.

Espelie KE, Sadek NZ, Kolattukudy PE. 1980. Composition of suberin-associated waxes from the subterranean storage organs of seven plants : Parsnip, carrot, rutabaga, turnip, red beet, sweet potato and potato. Planta 148: 468–476.

Fahn A. 1990. Plant Anatomy. 4th edn Pergamon press, Oxford

Garcin Y, Schwab VF, Gleixner G. 2012. Hydrogen isotope ratios of lacustrine sedimentary n- alkanes as proxies of tropical African hydrology: Insights from a calibration transect across Cameroon. Geochimica et Cosmochimica Acta 79:106–126.

Gocke M, Kuzyakov Y, Wiesenberg GLB. 2011. Differentiation of plant derived organic matter in soil, loess and rhizoliths based on n-alkane molecular proxies. Biogeochemistry 112: 23– 40.

Hilkert AW, Douthitt CB, Schlüter HJ, Brand WA. 1999. Isotope ratio monitoring gas chromatography/Mass spectrometry of D/H by high temperature conversion isotope ratio mass spectrometry. Rapid Commun Mass Spectrom 13: 1226–1230.

Hoffmann B, Kahmen A, Cernusak LA. 2013. Abundance and distribution of leaf wax n-alkanes in leaves of Acacia and Eucalyptus trees along a strong humidity gradient in northern Australia. Organic Geochemistry 62: 62–67.

Huang X, Wang C, Zhang J. 2011. Comparison of free lipid compositions between roots and Chapter 4: Organ δ2H values 125

leaves of plants in the Dajiuhu Peatland, central China. Geochemical Journal 45: 365– 373.

Jetter R, Kunst L, Samuels AL. 2006. Composition of Plant Cuticular Waxes. Blackwell Publishing Ltd, Oxford, UK, pp 145–181

Jones HG. 2014. Plants and microclimate: a quantitative approach to environmental plant physiology, 3th edn. Cambridge University Press, Cambridge, UK

Kahmen A, Schefuß E, Sachse D. 2013a. Leaf water deuterium enrichment shapes leaf wax n- alkane dD values of angiosperm plants I: Experimental evidence and mechanistic insights. Geochimica et Cosmochimica Acta 111: 39–49.

Kahmen A, Hoffmann B, Schefuß E. 2013b. Leaf water deuterium enrichment shapes leaf wax n- alkane δD values of angiosperm plants II: Observational evidence and global implications. Geochimica et Cosmochimica Acta 111: 50–63.

Körner C, Renhardt U. 1987. Dry matter partitioning and root length/leaf area ratios in herbaceous perennial plants with diverse altitudinal distribution. Oecologia 74: 411– 418.

Körner C. 1991. Some Often Overlooked Plant Characteristics as Determinants of Plant Growth: A Reconsideration. Functional Ecology 5:162.

Kuhn TK, Krull ES, Bowater A. 2010. Organic Geochemistry. Organic Geochemistry 41: 88–95.

Larcher W. 2003. Physiological Plant Ecology. 4th edn Springer Science & Business Media

Li Y, Beisson F, Ohlrogge J, Pollard M. 2007. Monoacylglycerols Are Components of Root Waxes and Can Be Produced in the Aerial Cuticle by Ectopic Expression of a Suberin-Associated Acyltransferase. Plant Physiology 144: 1267–1277.

Liu W, Yang H, Li L. 2006. Hydrogen isotopic compositions of n-alkanes from terrestrial plants correlate with their ecological life forms. Oecologia 150: 330–338.

Luo Y-H, Steinberg L, Suda S. 1991. Extremely Low D/H Ratios of Photoproduced Hydrogen by Cyanobacteria.

Peters K, Walters C, Moldowan J. 2005. The Biomarker Guide: Biomarkers and isotopes in the 126 Chapter 4: Organ δ2H values

environment and human history, 1 st edn Cambridge University Press , Cambridge, UK

Polissar PJ, Freeman KH. 2010. Effects of aridity and vegetation on plant-wax dD in modern lake sediments. Geochimica et Cosmochimica Acta 74: 5785–5797.

Potvin C. 1986. Biomass Allocation and Phenological Differences Among Southern and Northern Populations of the C4 Grass Echinochloa Crus-Galli. The Journal of Ecology 74: 915.

Retuerto R, Woodward FI. 1992. Springer for R&D. Oecologia

Rice KJ, Black RA, Radamaker G, Evans RD. 1992. Photosynthesis, Growth, and Biomass Allocation in Habitat Ecotypes of Cheatgrass (Bromus tectorum). Functional Ecology 6: 32.

Rickey MA, Anderson RC. 2004. Effects of nitrogen addition on the invasive grass Phragmites australis and a native competitor Spartina pectinata. Journal of Applied Ecology 41: 888– 896.

Rommerskirchen F, Plader A, Eglinton G. 2006. Chemotaxonomic significance of distribution and stable carbon isotopic composition of long-chain alkanes and alkan-1-ols in C4 grass waxes. Organic Geochemistry 37:1303–1332.

Sachse D, Radke J, Gleixner G. 2004. Hydrogen isotope ratios of recent lacustrine sedimentary n- alkanes record modern climate variability. Geochimica et Cosmochimica Acta 68: 4877– 4889.

Sachse D, Radke J, Gleixner G. 2006. δD values of individual n-alkanes from terrestrial plants along a climatic gradient – Implications for the sedimentary biomarker record. Organic Geochemistry 37:469–483.

Sachse D, Billault I, Bowen GJ. 2012. Molecular Paleohydrology: Interpreting the Hydrogen- Isotopic Composition of Lipid Biomarkers from Photosynthesizing Organisms. Annu Rev Earth Planet Sci 40: 221–249.

Salasoo I. 1989. Epicuticular wax alkanes of Ericaceae and Empetrum from alpine and sub-alpine heaths in Austria. Plant systematics and evolution 163: 71–79.

Sauer PE, Eglinton TI, Hayes JM. 2001. Compound-specific D/H ratios of lipid biomarkers from sediments as a proxy for environmental and climatic conditions. Geochimica et Chapter 4: Organ δ2H values 127

Cosmochimica Acta 65: 213–222.

Schläpfer B, Ryser P. 1996. Leaf and Root Turnover of Three Ecologically Contrasting Grass Species in Relation to Their Performance along a Productivity Gradient. Oikos 75:398.

Shaver GR, Billings WD. 1975. Root Production and Root Turnover in a Wet Tundra Ecosystem, Barrow, Alaska.

Schmidt H-L, Werner RA, Eisenreich W. 2003. Systematics of 2H patterns in natural compounds and its importance for the elucidation of biosynthetic pathways. Phytochemistry Reviews 2: 61–85.

Schwinning S, Starr BI, Ehleringer JR. 2005. Summer and winter drought in a cold desert ecosystem (Colorado Plateau) part II: effects on plant carbon assimilation and growth. Journal of Arid Environments 61: 61–78.

Sessions AL, Burgoyne TW, Schimmelmann A, Hayes JM. 1999. Fractionation of hydrogen isotopes in lipid biosynthesis. Organic Geochemistry 30: 1193–1200.

Shin M. 2004. How is Ferredoxin-NADP Reductase Involved in the NADP Photoreduction of Chloroplasts? Photosyn Res 80: 307–313.

Siegenthaler U, Oeschger H. 1980. Correlation of 18O in precipitation with temperature and altitude. Nature 285:314–317.

Smith FA, Freeman KH. 2006. Influence of physiology and climate on δD of leaf wax n-alkanes from C3 and C4 grasses. Geochimica et Cosmochimica Acta 70: 1172–1187.

Vogts A, Moossen H, Rommerskirchen F, Rullkötter J. 2009. Organic Geochemistry. Organic Geochemistry 40:1037–1054.

White D, Drummond J, Fuqua C. 2012. The physiology and biochemistry of prokaryotes, 4th edn. Oxford University Press, New York

Zhang X, Gillespie AL, Sessions AL. 2009. Large D/H variations in bacterial lipids reflect central metabolic pathways. Proc Natl Acad Sci USA 106: 12580–12586.

Zhang Y, Togamura Y, Otsuki K. 2004. Study on the n-alkane patterns in some grasses and factors affecting the n-alkane patterns. The Journal of Agricultural Science 142: 469–475. 128 Chapter 4: Organ δ2H values

Ziegler H, Osmond CB, Stichler W, Trimborn P. 1976. Hydrogen isotope discrimination in higher plants: Correlations with photosynthetic pathway and environment. Planta 128: 85–92. Conclusions 129

Chapter 5

Conclusions

The open questions addressed in the introduction and described throughout this thesis were answered as follows:

• Leaf water evaporative 2H-enrichment partially affects the leaf wax n-alkane δ2H

values in C3 and C4 grasses. This effect is not different between C3 and C4 grasses

(Chapter 2).

• Biosynthetic hydrogen fractionation (εbio) is significantly different between C3 and

C4 grasses (Chapter 2).

• After leaf maturity, n-alkane synthesis does not cease. However, the secondary

synthesis is relatively low and varies across C3 grasses (Chapter 3).

• Besides leaves, other grass organs synthesize n-alkanes. Carbon autonomous and

non-carbon autonomous organs have significantly different n-alkane δ2H values in

C3 grasses (Chapter 4).

5.1. Discussion

5.1.1. n-Alkane concentration in grasses in relation with the cuticular wax function

The biological function of plant cuticular lipids such as n-alkanes has been classically theorized to protect the plant from natural stressors, environmental abrasion and non- stomatal water losses (Jetter and Schäffer, 2001; Jetter et al., 2006). This has been however not directly tested. During the conduction of two experiments (Chapters 2 and 3) the grasses 130 Acknowledgments

that were grown under wet conditions (75% RH) had lower n-alkane concentrations than grasses grown under dry (50% RH) conditions. Other studies have described a tendency to have lower leaf wax n-alkane concentrations in humid than in dry conditions for woody plants (Sachse et al., 2006; Hoffmann et al., 2013). It is also possible that the larger VPD demand experienced in low humidity affects the cuticular lipid concentration. Among grass organs, inflorescences had the largest concentration of n-alkanes (Chapter 4). Interestingly they also had important quantities of short-chain alkanes, which are usually rarely present in grass leaves. Inflorescences are situated higher in the grassland canopy where they are exposed to large evaporative demand and wind abrasion. The larger concentration and particular distribution of n-alkanes in inflorescences could be thus an evolutionary strategy to avoid desiccation while ensuring reproductive success. The increased n-alkane concentration in inflorescences in response to low humidity suggests therefore, as in the case of humidity treatment discussed above (Chapters 2 and 3), a dependency of the wax n- alkane concentration on plant-water conditions. In natural environments it was also found, that the leaf wax n-alkane concentration was generally higher in grasses at Alp Weissenstein

(alpine grassland) than at Ennetbaden (temperate grassland) (Chapter 4). This could be associated with environmental conditions in high altitudes (Dodd and Poveda, 2003). For example at high altitudes a cuticular thickness dependency on UV radiation has been indicated (Anfodillo et al., 2002).

The impact of inflorescence-derived n-alkanes on the sedimentary record depends not only on concentration but also on the overall production of inflorescences, turnover and transport processes occurring in a growth season. It is difficult to give a single value to express the contribution of inflorescences to the sedimentary records because such processes are highly dependent on multiple species-specific conditions and environmental variables. In a simple mass balance calculation it was assessed that leaves are largely the main contributor of n-alkanes (more than 83%) in the soils and sediment matrixes and Conclusions 131

therefore it is suitable to use leaf material in paleohydrological or plant physiological research.

5.1.2. Effect of leaf water evaporative 2H-enrichement on leaf wax n-alkane δ 2H values

Plant leaf water isotope ratios (δ2H and δ18O values) have been widely studied and modeled.

Their isotope composition is affected by multiple environmental and leaf physiological factors (Craig and Gordon, 1965; Dongmann et al., 1974; Kahmen et al., 2011b). In some cases it is unclear what the drivers of across species variability are. It has been found for example that C4 grasses have larger evaporative 18O-enrichment than C3 grasses (Helliker and Ehleringer, 2000). It was hypothesized that anatomical differences such as shorter intervenial length in C4 plants could drive these differences. A shorter intervenial length would enhance the mixing of the biosynthetic water with isotopic enriched water from the evaporative sites. An additional explanation is that compared to C3 plants, C4 plants are ecologically adapted to more arid conditions. C4 grasses are therefore prone to down- regulate transpiration upon environmental evaporative demand. This would, according to the Péclet effect in water isotope theory, result in a greater evaporative enrichment in C4 plants. In contrast, this study has shown that leaf water evaporative 2H-enrichment (LW Δ2H) was identical for C3 and C4 grasses (Chapter 2). Consequently, different n-alkane δ2H values between C3 and C4 grasses are not explained by distinctive LW Δ2H. As such, different n- alkane δ2H values between C3 and C4 grasses are therefore not induced by a different anatomy, a property that is often described as a major isotopic control.

A partial effect of the LW Δ2H on the n-alkane δ2H values was found. This effect varied across species and did not differ between C3 and C4 grasses (Chapter 2). On average, the effect of LW Δ2H on n-alkane δ2H values was 64% and 39% for C3 and C4 grasses respectively. It is however difficult to assign a single value to this effect in C3 as well as in C4 grasses. This is because such an effect ranged from 1% to 100%. Nevertheless, our results 132 Acknowledgments

are in full agreement with evidence from other studies that also suggest a partial dependency of n-alkane δ2H values on LW Δ2H (Feakins and Sessions, 2010; Kahmen et al.,

2013a) and for the first time, this effect was quantified for a wide range of grass species. The large variability across C3 and C4 species found in this study could be explained by the species-specific drivers of the isotopic status of the biosynthetic water pool at the intercalary meristem, where the primary sites of leaf wax formation are. In contrast to monocots, dicots transfer 100% of the LW Δ2H to the n-alkane δ2H values (Kahmen et al.,

2013a). The different LW Δ2H contribution to the n-alkane δ2H values between monocots and dicots has been explained by differences in leaf growth and development between these two plant groups. It was also investigated what the potential drivers of the heterogeneity of the effect of LW Δ2H on n-alkane δ2H values across C3 species are, e.g. plant transpiration (Chapter 2). A poor correlation between such an effect and transpiration rates in C3 grasses was found. Consequently, there is no n-alkane δ2H values dependency on transpiration. At least transpiration alone does not have an impact on the contribution of

LW Δ2H to the biosynthetic water pool at the base of the meristem in grasses.

These results have important implications for paleoecology. For example, typically lower leaf wax n-alkane δ2H values in monocots (compared to dicots) would be driven mainly by an incomplete transfer of the leaf water enrichment to the biosynthetic pool. As such, the n- alkane isotopic composition from grasses will be generally less 2H-enriched than from woody plants.

5.1.3. Low secondary n-alkane synthesis in grasses

It was found that there is a secondary n-alkane synthesis after leaf maturation in C3 grasses

(Chapter 3). The secondary n-alkane synthesis rates were low and ranged between 0.09 to

1.09% per day across C3 grass species. Those results are in agreement with previously published data for grasses (Gao et al., 2011) but contradict findings of secondary n-alkane Conclusions 133

synthesis for dicots, which were found to be not existing or at least negligible (Kahmen et al.,

2011a; Gao et al., 2012). The secondary n-alkane synthesis found here is relatively low compared to the majority of n-alkane synthesis at the intercalary meristem at the base of the blade that occurs at the beginning of the life of a leave. However it is possible that larger secondary synthesis of lipids would occur in natural conditions where a mix of environmental stressors could trigger a new cuticular n-alkane synthesis. This is particularly feasible in open grasslands where the growth season is long and not only low humidity but also other environmental factors play a role, e.g. intense wind, high UV light or periods of extreme drought. Additionally it was also hypothesized (Chapter 2) that secondary n-alkane synthesis after leaf maturation would explain the variability in the effect of LW Δ2H on n- alkane δ2H values. n-Alkane synthesis in a mature leave would not depend on the conventional biosynthetic water pool at the base of the meristem but on 2H-enriched water from the evaporative sites. This is however not the case since the secondary n-alkane synthesis was low in C3 grasses. If transpiration and secondary n-alkanes synthesis do not explain the heterogeneity of the effect of LW Δ2H on n-alkane δ2H values, then there has to be a mix of other factors e.g. anatomical traits, physiological factors or biochemical variables, affecting the biosynthetic water pool.

This has implications for environmental and physiological studies. Literature has shown both, constant and variable n-alkane δ2H values in different plants during a growth season (Pedentchouk et al., 2008; Sachse et al., 2009; Sachse et al., 2010; Feakins and

Sessions, 2010; Tipple et al., 2013; Sachse et al., 2015). This has been explained by a potential secondary synthesis of n-alkanes in different plant types. The results presented here show that a continuous regeneration or secondary n-alkane synthesis in grasses would in any case be only partially detected because n-alkane δ2H heterogeneity would be mainly caused by environmental or plant physiological conditions affecting the synthesis of n- alkanes at the beginning of the season. As such, low secondary n-alkane synthesis implies no 134 Acknowledgments

significant repercussion of environmental or plant physiological conditions occurring after leaf maturity on grass-derived leaf wax n-alkane δ2H values in geological records.

5.1.4. Variability of the biosynthetic fractionation and its implications

2 Most interestingly, the biosynthetic hydrogen isotope fractionation (εbio) on n-alkane δ H values was significantly different between C3 and C4 grasses (Chapter 2). This is in full

agreement with previously published data, which shows more negative εbio, i.e. a larger fractionation in the synthesis of n-alkane for C3 than for C4 grasses (Smith & Freeman, 2006;

McInerney et al., 2011). εbio depends on several fractionation processes. The NADPH-derived hydrogen sources play also an important role during the biosynthesis of n-alkanes. NADPH from the light reaction of photosynthesis is more depleted in deuterium than NADPH from the pentose pathway of stored carbohydrates. It is hypothesized that C3 and C4 grasses might be using these hydrogen sources in different proportions. Recent work in bacteria showed a dependency of lipid isotope composition on carbon autonomy (Zhang et al., 2009).

Following this evidence, C3 grasses would make more use of recent assimilates than C4 grasses, which would rather use larger amounts stored carbohydrates for their physiological processes. Similarly than for C3 and C4 types, it was found that non-carbon autonomous organs such as inflorescences and roots have higher n-alkane δ2H values than autonomous carbon organs such as leaves (Chapter 4). This means that organ-derived n-alkane δ2H values are determined by the origin of hydrogen during compound synthesis. Thus it is likely that non-carbon autonomous organs, such as inflorescences and roots, acquire hydrogen from storing compartments while autonomous carbon organs, such as leaves, generate their hydrogen from photosynthesis.

The biochemical processes investigated here for grasses have important implications for paleoecological studies. Generally speaking the different in n-alkane δ2H values between

C3 and C4 grasses might not be driven by LW Δ2H but rather be the result of systematic Conclusions 135

differences in εbio between these two plant groups. As such, ecosystem changes, e.g. C3 to

C4 biomes, would affect more the sedimentary record and its δ2H values than other hydrological changes.

5.2. Outlook

Given the major findings presented here, the following are three main areas that future research could address:

• Ecological function of plant cuticular lipids

Even though this thesis did not focus on the study of the ecological function of

cuticular wax lipids, the data presented here suggest that a high evaporative

demand and a high light radiation would exert the synthesis of cuticular n-alkanes in

order to decrease evaporative water loss from the cuticle. An area of future research

could be therefore the understanding of the role of cuticular waxes and the different

controls of lipid concentrations and distributions i.e. drought and light. This would

provide a better understanding of plant-derived cuticular lipids in ecology and in

their applications as environmental biomarkers.

• Biosynthetic water pool in grasses

This thesis found a variable effect of LW Δ2H on leaf wax n-alkane δ2H values in C3

and C4 grasses. Two potential drivers of this variable effect were further

investigated, i.e. transpiration and secondary n-alkane synthesis. However, none of

them seem to explain the large variability of such an effect. This indicates that other

anatomical or physiological processes might be controlling the isotopic signature of

the biosynthetic water pool. Thus, studies could focus on investigating anatomical

differences such as mesophyll and xylem ratios, intervenial length or 136 Acknowledgments

compartmentalization in relation with the leaf biosynthetic water pool. This would

improve our understanding of leaf physiology and compound biosynthesis.

• Hydrogen sourcing connected to carbon internal use of plants

Evidence from this thesis and previous studies suggests links between plant n-alkane

δ2H values and NADPH-derived hydrogen source. The H-NADPH originates either

from the light reaction of photosynthesis or from or from oxidized carbohydrates. As

such, there is a dependency of the δ2H values from specific compounds on plant-

carbon relationships. There is, however, an incomplete understanding of the links

between hydrogen in n-alkanes and carbon relationships. Therefore, new studies

should aim at a better understanding of the biochemical sourcing of hydrogen in

lipids and also, the environmental variables that could affect such hydrogen source.

This is exciting n-alkane δ2H values could provide insights into the biochemical

carbon autonomy and energy metabolism. Moreover n-alkane δ2H values could be

linked to the plant carbon internal use in a context of environmental global change.

Acknowledgments 137

Acknowledgements

It was a long and exciting way. I only have words of appreciation to all those people that in different way contributed, helped and also enjoyed with me during this time:

Thanks to Ansgar who gave me the chance to start in this fascinating area of plant physiology, also for all his terrific mentoring even when in Basel his constant support motivated and could put me back in a good track when needed (which is not always easy).

To Nina who provided me with a foster researching home when I needed it, making me feel at home and even when she was not my direct supervisor it was great to learn so much from her.

Many thanks to Tim who let me be in his lab and interact with great people of his team. He also accepted to be my examiner three times with no problem.

Rolf is who initiated on me an interest for plant research, back in the days, I had the incredible chance to meet him and work with him at PSI.

To all this fantastic grassland group. Nicole, full smiles always, always, ready to help in any admin trouble. Werner always with a story to tell was very nice having him around in the lab. Carmen, Petra, Sebastian, Lutz, Lukas giving a tender and supportive hand. Peter,

Thomas, Phillip, Florian and Patrick, always the workshop with open doors to help out in anything. Annika and Roland are responsible of literally thousands of analyzed samples in the IRMS, weekends, late evenings, unstoppable, thanks guys.

To all these GL-PhD army who was there to help, talk or have a beer. Nadine was the first office roommate I had, not so much work together we had tons of fun in the office and she always was there to carry on with me in the PhD. Thanks Nady. Marc-André my old Nemesis, what topic has not been discussed with you, from waxes to Greek philosophy. Thanks for the 138 Acknowledgments

great scientific partnership. Carola, Dörte, Marco and Sämi were always there to share a good moment and more recently Neringa, Shiva, Elena, Eugenie and Thomas who I tried to pass on my outside-of-PhD knowledge not very successfully tho. To Maura, Günter, Sarah,

Dan, Claudia, Lars, Rafi and Victor for that support in Basel.

Special moments shared with my second family from altstetten: David, Christian, Juan, Raj,

Krzys and Andy. Thanks boys.

A Julito, gracias wy por ser mi otro hermanito menor, la he pasado muy bien contigo y espero que sigamos siempre asi.

Franco, meshi ya tu sa.

Last but not least, to Christine who might be reading all these pages first than anybody as she proof read all. Thanks Loquinena for being there.