CONSTRAINTS OF MICROCLIMATE AND THE ALPINE … · stressful conditions for the focal organisms. Alpine treelines are among the most conspicuous vegetation limits in existence, and

THE REGENERATION NICHE OF TREES AT THE ALPINE TREELINE

-

CONSTRAINTS OF MICROCLIMATE AND THE ALPINE GRASSLAND

VEGETATION ON GERMINATION AND SEEDLING ESTABLISHMENT

Die von der Fakultät für Mathematik und Naturwissenschaften

der Carl von Ossietzky Universität Oldenburg

zur Erlangung des Grades und Titels eines

Doktors der Biologie (Dr. rer. nat.)

angenommene Dissertation von

Hannah Loranger, geboren am 26.11.1985 in Duisburg

Erstgutachter: Prof. Dr. Gerhard Zotz

Zweitgutachter: Prof. Dr. Martin Diekmann

Tag der Disputation: 30.05.2016

v

CONTENTS

Chapter 1 – General Introduction ....................................................................................... 7

Defining the term “treeline”............................................................................................... 8

Historical retrospect and recent interest .......................................................................... 9

Causes of treeline formation ............................................................................................ 11

The regeneration niche of treeline trees ......................................................................... 13

Thesis outline...................................................................................................................... 15

Chapter 2 – Impacts of soil microclimate on early establishment of trees at the

alpine treeline: idiosyncratic responses and the importance of soil moisture ......... 19

Abstract ............................................................................................................................... 19

Introduction ........................................................................................................................ 21

Methods ............................................................................................................................... 25

Results ................................................................................................................................. 32

Discussion ........................................................................................................................... 39

Consistency of limiting factors during early establishment ................................... 39

Temperature, moisture, and their interactions driving early establishment

success ............................................................................................................................. 40

Species-specific responses ............................................................................................ 42

Implications for local treeline patterns and dynamics ............................................. 46

Conclusions ........................................................................................................................ 48

Acknowledgements ........................................................................................................... 48

Appendix ............................................................................................................................ 50

Chapter 3 – Competitor or facilitator? The role of grassland vegetation for

germination and seedling performance of tree species at the alpine treeline ......... 53

Abstract ............................................................................................................................... 53

Introduction ........................................................................................................................ 56

Materials and Methods ..................................................................................................... 59

Results ................................................................................................................................. 66

Discussion ........................................................................................................................... 74

Competition dominates the interaction between tree seedlings and neighbouring

vegetation ....................................................................................................................... 75

vi

Seasonal shifts from competition to facilitation depend on the leaf functional

type .................................................................................................................................. 79

Acknowledgements ........................................................................................................... 82

Appendix ............................................................................................................................ 83

Chapter 4 – A cool experimental approach to explain elevational treelines, but can

it explain them? ..................................................................................................................... 87

Abstract ............................................................................................................................... 87

Introduction ........................................................................................................................ 89

Confounding temperature conditions ............................................................................ 91

Confounding moisture conditions .................................................................................. 93

Alternative experimental setups ..................................................................................... 95

Alternative treeline-forming mechanisms in Nothofagus and other genera .............. 99

Conclusion ........................................................................................................................ 102

Appendix .......................................................................................................................... 103

Chapter 5 – Synthesis ........................................................................................................ 105

Local and intrinsic factors driving the regeneration response .................................. 106

Outlook and future research needs ............................................................................... 110

References ............................................................................................................................ 114

Summary .............................................................................................................................. 132

Zusammenfassung ............................................................................................................. 134

Danksagung ......................................................................................................................... 137

Lebenslauf ........................................................................................................................... 139

Authors’ contributions ...................................................................................................... 141

Allgemeine Erklärung ....................................................................................................... 143

7

CHAPTER 1

GENERAL INTRODUCTION

Life on earth is incredibly diverse, but none of the millions of species already

catalogued or still to discover is evenly distributed over the available space. The

arising fundamental questions of “where?”, “how?” and “why?” encompass the

central interest of ecology: to describe and understand the interactions that

determine the distribution and abundance of organisms (Krebs 1972). In this context,

a special relevance can be attributed to distributional boundaries, which offer unique

opportunities to study limiting factors and their interactions due to increasingly

stressful conditions for the focal organisms. Alpine treelines are among the most

conspicuous vegetation limits in existence, and mark not only the distribution

boundary of single tree species, but of the entire life-form “tree”, due to an

increasingly unfavourable heat balance (Körner 2012). They are at the same time

important ecological boundaries, since site conditions such as topoclimate and soil

Chapter 1

8

properties change significantly over the transition from closed forest to alpine tundra

(Holtmeier 2009). These characteristics designate alpine treelines as particularly

interesting study systems, which, although having been linked to a common

isotherm of 5 – 7 °C mean growing season temperature on a global scale (Körner and

Paulsen 2004), still lack comprehensive mechanistic explanations for locally varying

patterns and elevational positions. This thesis aims at contributing to a more

complete understanding of alpine treeline patterns and dynamics by focussing on the

early establishment of treeline trees with regard to abiotic as well as biotic potentially

limiting factors.

DEFINING THE TERM “TREELINE”

There is no general consensus reached yet regarding the definition and application of

the term “treeline”, clearly distinguishing it from similar boundary concepts such as

“timberline” or “forest line” (Körner 1998). For this reason I adopted the concise and

suitable definition established by Bader (2007), which uses treeline as an abbreviation

for treeline ecotone, as it is rather a transition zone than a sharply delimited line (Fig.

1.1). Treeline is then defined as “the existent transition between continuous upper

montane forest and continuous alpine vegetation”. Therein, the term “existent” refers

to the actual presence of a treeline, irrespective if it is at the climatic limit or

depressed by anthropogenic influences. The term “continuous” emphasizes the

“clear dominance of either vegetation type”.

General Introduction

9

Fig. 1.1 Relatively undisturbed treeline of the Rif Blanc Forest, Galibier massif,

French Alps at approximately 2300 m a.s.l. adjoining the subalpine pastures of the

historically deforested Lautaret Pass. The dominant tree species are Larix decidua and

Pinus uncinata; land-use impacts are limited to occasional sheep-grazing. A. From a

distant view the limit of tree cover appears to follow a relatively sharp line. B. A

close-up of the same treeline reveals its transitional nature, with reduced tree size

and increasing patchiness of tree cover with increasing elevation.

HISTORICAL RETROSPECT AND RECENT INTEREST

The elevation of tree cover on high mountains was essentially modified by human

activities as early as the Bronze Age (Wick and Tinner 1997), however, it took a long

time in the course of history until a scientific interest in this noticeable vegetation

boundary arose. The first descriptive records of elevational vegetation belts date to

observations of the renaissance polymath Leonardo da Vinci (1452-1519) in the

Italian Alps, followed by those of Conrad Gessner (1515-1565) in the Swiss Alps. The

latter was presumably the first to mention an elevational limit of tree growth as

related to climatic factors, namely temperature and growing season length

(Holtmeier 1965). In the following centuries, remarks on alpine treelines were rather

accidental in connection with geological or botanical studies, and only about 200

A B

Chapter 1

10

years ago this vegetation boundary itself was moved more into the focus of scientific

inquisitiveness (Marek 1910). An impressive prelude was given by the famous Latin

America – expedition of Alexander von Humboldt (1769-1859), from which he laid

the foundations of biogeography by relating elevation and plant distribution on

mountain slopes to isotherms, using treelines as an important reference point (von

Humboldt and Bonpland 1805). The initiation of actual treeline research started with

systematic measurements of treeline elevations (e.g. Wahlenberg 1813) and, as a

consequence of previous observations, the study of different thermal parameters as

potentially overarching explanation (reviewed in Holtmeier 2009). The results of this

early research were synthesized in the extensive works of Brockmann-Jerosch (1919)

and Däniker (1923), the former considering the complex climate character of treelines

on a global scale and the latter developing a thorough theory about tree growth

limitations at high elevations due to heat deficiency that is still in the centre of

debates today. Däniker was also the first to apply ecological research methods, which

were extended in the 1930s with a stronger focus on tree physiology by several

pioneers of experimental ecology (e.g. Michaelis 1934; Steiner 1935; Pisek and

Cartellieri 1939). The heavy destruction caused by avalanches in the European Alps

in the 1950s then gave rise to a new perspective on alpine treelines with regard to

social and economic value of ecosystem services, stimulating more studies on the

potential limit of protective mountain forests and motivating the establishment of

specific research stations (Obergurgl 1953, Austria; Stillberg/Davos 1959,

Switzerland; Holtmeier 2009). This increase of local or regional, often species-specific

knowledge coincided with an increasingly broader approach of treeline research,

including large-scale bioclimatological works (Daubenmire 1954; Hermes 1955) and


11

worldwide comparisons of alpine treelines (Troll 1973; Holtmeier 1974; Wardle 1974),

highlighting the global nature of the treeline phenomenon. From these currents of

treeline research emerged two distinct approaches with differing focuses of interest

that dominate the field today: while the global approach aims at determining the

general factors causing the formation of treelines, the local or landscape-approach

seeks to understand why treelines differ among locations (Malanson et al. 2011).

The rapidly increasing number of publications in both lines of research in

recent years is the consequence of a new global challenge, climate change. Global

temperatures have risen by about 0.7 °C over the last century and climate models

predict a further increase of 1-5 °C for the century to come (IPCC 2013). As a

thermally limited distribution boundary treelines are expected to move upslope or

poleward (Grace 2002), and indeed, there is evidence for advancing tree cover to

higher elevations and latitudes as well as increased radial growth in treeline trees

(Rolland et al. 1998; Kullman 2007; Shiyatov et al. 2007; Vittoz et al. 2008). However,

responses are not uniform, since locally stable or even receding treelines have been

reported as well (Harsch et al. 2009). In this regard, the recent interest in the treeline

topic reflects the urgent need of a more mechanistic understanding of limiting factors

at different spatial scales, which is a prerequisite for reliable predictions of treeline

shifts and the associated consequences for sensitive mountain ecosystems.

CAUSES OF TREELINE FORMATION

On a global scale, the elevational limit of trees is ultimately set by heat deficiency.

While this relationship was observed early in the history of treeline research, it was

especially the establishment of the growth-limitation hypothesis postulated by

Chapter 1

12

Körner (1998), which connected correlations of isotherms and global treeline

positions with a likely mechanism. This hypothesis comprises that the life-form tree

is characterized by an upright stem with high-reaching branches, which are tightly

coupled to the atmosphere and its temperature fluctuations. The resulting low tissue

temperatures under harsh conditions limit cell division and thus growth,

distinguishing trees from low-stature alpine vegetation with a much more favourable

heat balance. This concept is supported to the disadvantage of competing

explanations such as the carbon-limitation hypothesis (Stevens and Fox 1991) by

strong evidence: i) at the lower temperature threshold of growth of approximately 5

°C photosynthesis rates are still positive (Grace 2002; Rossi et al. 2008), ii) this lower

limit of growth corresponds well to the isotherm of 5 – 7 °C associated to positions of

treelines worldwide (Körner and Paulsen 2004), which can differ however

substantially in more stress-related factors such as seasonality, snow cover or solar

irradiance (e.g. comparing temperate and tropical alpine treelines) and iii) many

studies report an increase of mobile carbon stores with increasing elevation,

suggesting that while the assimilation of photosynthates is not restricted at treeline,

their utilisation for growth might be (e.g. Hoch and Körner 2003, 2012; Shi et al. 2008;

Fajardo et al. 2012).

However, limitations of carbon gain as well as a set of other hypothesis

regarding the causes of treeline formation including climatic stress, nutrient

deficiency and disturbances (summarized in e.g. Tranquillini 1979; Körner 1998;

Wieser and Tausz 2007) may well play an important role at a local scale. As i)

treelines can vary locally up to several hundred metres in their elevational positions,

ii) patterns range from abrupt treelines over scattered tree islands to diffuse forms


13

and iii) treelines do not all respond uniformly to a warming climate, locally varying

environmental factors unrelated to temperature must be responsible (Holtmeier and

Broll 2005; Harsch and Bader 2011; Case and Duncan 2014). Among the abiotic

conditions especially water availability seems to play a critical role, as moisture

deficits can be positively linked to climate warming and have the potential to

override positive treeline responses to increasing temperatures (e.g. Barber et al.

2000; Daniels and Veblen 2004). Further site-specific complexity is added by biotic

factors, because interactions with other organisms can be beneficial (Mattes 1994;

Akhalkatsi et al. 2006) or detrimental (Moir et al. 1999; Cairns and Moen 2004) for

treeline trees. These abiotic and biotic site-specific conditions, but also intrinsic

factors such as tree species ecology (Körner and Paulsen 2004) and life-stage

dependent requirements (Barbeito et al. 2012) may interact with temperature to

determine local treeline patterns and dynamics. Thus, studying these

interdependencies is essential for a more complete understanding of the mechanisms

governing alpine treelines, thereby exploring what prevents some treelines from

reaching their climatic limit and facilitating predictions of future dynamics.

THE REGENERATION NICHE OF TREELINE TREES

The regeneration niche of a species can be defined as “the expression of the

requirements for a high chance of success in the replacement of one mature

individual by another” (Grubb 1977). For treelines to remain stable or to advance

even to higher elevations, such a replacement of trees within and above this ecotone

is a necessary prerequisite. However, the regeneration niche of trees can be expected

to become constricted under the increasingly harsh conditions and the substantially

Chapter 1

14

different ecosystem properties in the alpine tundra (Fig. 1.2A). While seed

availability and successful dispersal are important precursors, germination and

seedling establishment are the critical steps that determine the compatibility between

the environment and the respective regeneration niche, and are widely recognized as

major life-history bottlenecks for treeline trees (Stevens and Fox 1991; Smith et al.

2003). Germination depends on environmental cues such as a specific temperature or

moisture range and is the first, critical life-stage transition (Baskin and Baskin 2001).

Subsequently, young seedlings are in the most vulnerable life-stage of a tree due to

their small size and the resulting low water and reserve storage capacity (Cui and

Smith 1991; Wang and Zwiazek 1999; Johnson et al. 2011). Furthermore, a different

physiology compared to adult trees (e.g. Tegischer et al. 2002) might contribute to a

varying or higher sensitivity to environmental stressors (Germino et al. 2002;

Barbeito et al. 2012; Moyes et al. 2013). These characteristics are reflected in the

natural distribution of seedlings in the treeline ecotone, which often shows a

decreasing density along the elevational gradient (Cuevas 2000) and a spatial

association with shelter elements (Germino and Smith 1999; Smith et al. 2003; Batllori

and Camarero 2009). In this context, other plants such as those forming alpine

grassland vegetation might play an ambiguous role for tree regeneration (Fig. 1.2B),

as they can not only provide shelter by improving microclimatic conditions (Smith et

al. 2003; Maher et al. 2005), but also act as strong competitors for resources (Nambiar

1990; Moir et al. 1999). Also the specific types of treelines, especially abrupt forms,

appear tightly linked to regeneration patterns (Harsch and Bader 2011), and might

arise from a lack of safe sites (sensu Harper 1977) in the alpine tundra (Tranquillini

1979). Consequently, understanding the limitations of tree recruitment at their


15

distribution boundary is a key component in the discussion about treeline causality

and predictions.

Fig. 1.2 A. Young Sorbus aucuparia - seedling experiencing stressful high-light

conditions on a sunny morning following a frost night. B. Freshly germinated Larix

decidua – seedling surrounded by dense herbaceous alpine vegetation, which might

act as strong competitor or as facilitator providing shelter.

THESIS OUTLINE

The aim of this thesis is to deepen the knowledge of which environmental factors,

considering abiotic and biotic interactions, constrict the regeneration niche of treeline

trees, and how. To include the potentially most important drivers of local treeline

patterns (i.e. site-specific environmental factors, life-stage-specific responses, tree

species ecology), I investigated the effect of abiotic and biotic interactions with

respect to i) the two critical, successive first life-stages germination and early

seedling establishment and ii) general or species-specific responses by comparing

five important European treeline tree species. Hereafter I will give a brief outline of

the specific focus of the following chapters 2 to 5.

Chapter 2 and 3 present results obtained in common garden field experiments

conducted in the experimental garden of the alpine research station Joseph Fourier in

the French Alps near the local treeline (Lautaret Pass, 2100 m a.s.l., Fig. 1.3). The

A B

Chapter 1

16

possibility of combining the experimental manipulation of the microclimate as well

as biotic interactions with fine-scale measurements of two critical early life-stages in

multiple species produced a unique dataset, addressing for the first time in unity this

complexity of interdependent factors and their potential to shape local treelines.

In Chapter 2 we tested the impact of microclimate – more specifically the

effects of temperature, water availability and the interaction of both – on germination

and first-year seedling survival of the focal species. Besides the well-established

importance of temperature, water availability was shown to affect treelines in general

(González de Andrés et al. 2015) and their regeneration in particular (Moyes et al.

2013). As both climate variables can be strongly interdependent, their interaction

might be more important for treeline responses than their absolute values (Ohse et al.

2012).

Chapter 3 deals with the ambiguous role of alpine grassland vegetation for

tree establishment at and above the alpine treeline, since both positive (Germino et

al. 2002; Maher et al. 2005) and negative (Moir et al. 1999; Dullinger et al. 2003)

interactions have been documented before. Here, we focussed on stress- (survival) as

well as resource-related responses (biomass, mobile carbon reserves) to establish the

relative importance of facilitative and competitive effects.

Additionally, we planned a field experiment to address the effect of the

changing microclimatic conditions for the growing tree seedling as it emerges from

the buffered vegetation cover and becomes increasingly coupled to the atmosphere,

since this important transition might account for life-stage-specific susceptibilities to

environmental stressors (Barbeito et al. 2012) and growth limitations with increasing


17

size (Körner 1998). To test the effect of seedling height within and above the alpine

grassland vegetation independently from ontogenetic influences and tree size, we

conceived planting tubes that were vertically adjustable on a metal pole, in which

seedlings of the same age class could be grown at different heights within and above

the vegetation layer. However, preliminary temperature measurements in the

planting tubes in the ground and suspended at 1 m height showed that in spite of

insulation, ventilation layers and reflective shields, soil temperatures were not

comparable and would thus bias the seedling response to the treatments.

Nevertheless, this testing provided valuable experience and data, which we applied

in Chapter 4 to contest a study (Fajardo and Piper 2014) using a similar design

without controlling neither seedling root-zone temperature nor soil moisture. In this

chapter, the consequently questionable interpretation of results in Fajardo and

Piper’s study is put in context with the potential impact of fluctuating temperatures

and moisture deficits, and alternative experimental setups for this after all elegant

approach are explored.

Finally, Chapter 5 synthesises the results and conclusions of the previous

chapters and gives an outlook to future research.

Chapter 1

18

Fig. 1.3 View from the Chaillol Mountain on the Lautaret Pass (2050 m a.s.l.),

showing the Alpine Research Station Joseph Fourier of the University Grenoble with

the associated Lautaret Alpine Botanical Garden (white arrow) and the experimental

garden (black arrow, 2100 m a.s.l.). On the left in the background, relatively

undisturbed treelines principally formed by Larix decidua and Pinus uncinata can be

seen at approximately the same elevation.

19

CHAPTER 2

IMPACTS OF SOIL MICROCLIMATE ON EARLY ESTABLISHMENT OF

TREES AT THE ALPINE TREELINE: IDIOSYNCRATIC RESPONSES AND

THE IMPORTANCE OF SOIL MOISTURE

Hannah Loranger, Gerhard Zotz, Maaike Y. Bader

Submitted

ABSTRACT

On a global scale, temperature is the main determinant of arctic and alpine treeline

position. However on a local scale, treeline form and position vary considerably due

to other climatic factors, tree species ecology and life-stage-dependent responses. For

treelines to advance poleward or uphill, the first steps are germination and seedling

establishment. These earliest life stages may be major bottlenecks for treeline tree

populations and will depend differently on climatic conditions than adult trees. We

Chapter 2

20

investigated the effect of soil temperature and moisture on germination and early

seedling survival in a field experiment in the French Alps near the local treeline (2100

m a.s.l.) using passive temperature manipulations and two watering regimes. Five

European treeline tree species were studied: Larix decidua, Picea abies, Pinus cembra,

Pinus uncinata and Sorbus aucuparia. In addition, we monitored the germination

response of three of these species to low temperatures under controlled conditions in

growth chambers. The early establishment of these trees at the alpine treeline was

limited either by temperature or by moisture, the sensitivity to one factor often

depending on the intensity of the other. The results showed that the relative

importance of the two factors and the direction of the effects are highly species-

specific, while both factors tend to have consistent effects on both germination and

early seedling survival within each species. We show that temperature and water

availability are both important contributors to establishment patterns of treeline trees

and hence to species-specific forms and positions of alpine treelines. The observed

idiosyncratic species responses highlight the need for studies including several

species and life-stages to create predictive power concerning future treeline

dynamics.

Keywords: alpine treelines, climate change, early seedling survival, germination,

temperature-moisture-interactions, time-to-event-analysis

Effects of Microclimate

21

INTRODUCTION

Treelines are conspicuous transition zones between two very different vegetation

types. There is a growing concern about how global climate change may affect these

systems, and as a consequence much attention has been drawn to both alpine and

arctic ecotones in recent years. Treelines could represent a distinct indicator of

climate warming since temperature is recognized as the main determinant of treeline

position on a global scale, roughly following a common isotherm of 5-7 °C mean

growing season temperature (Körner and Paulsen 2004). Many studies show recent

advances of treelines poleward and to higher elevations, as well as increasing radial

growth of the trees forming these ecotones (Rolland et al. 1998; Kullman 2007;

Shiyatov et al. 2007; Qi et al. 2015). However, stable or receding treelines have been

found (Harsch et al. 2009), and treeline position may vary considerably at a local

scale (Holtmeier and Broll 2005; Case and Duncan 2014). Such local variations can be

due to locally varying environmental conditions unrelated to temperature such as

precipitation (Holtmeier and Broll 2005), tree species ecology (Körner and Paulsen

2004) and life-stage dependent environmental dependencies (Barbeito et al. 2012;

Greenwood et al. 2015).

These abiotic and biotic factors may also interact with temperature to

determine the form and dynamics of a treeline at a specific site. For example, the

consequences of moisture deficits – which can be positively linked to climate

warming – have been shown to override positive temperature responses with respect

to growth (Barber et al. 2000; González de Andrés et al. 2015) and regeneration

(Barton 1993; Daniels and Veblen 2004; Moyes et al. 2013). In such cases, treeline

Chapter 2

22

shifts may depend more on the interactions of temperature and water availability

than on their absolute values (Ohse et al. 2012). As a result, it is commonly observed

that tree cover is slow or unable to expand to its ultimate thermal boundary

(Holtmeier 2009). The underlying mechanisms remain however difficult to

disentangle and there is an urgent need for quantitative assessments of the specific

environmental conditions and associated mechanisms preventing the establishment

of different tree species beyond current treelines.

Treelines represent distribution boundaries for an entire life-form – the tree.

Consequently, ecosystems above the treeline differ fundamentally from those below,

e.g. in regard to soils and microclimates (Sullivan and Sveinbjörnsson 2010; Thébault

et al. 2014). This presents particular challenges for a successful tree regeneration and

establishment in the treeline ecotone and beyond, as required for an upward

distributional shift. Previous studies have shown that traits essential for regeneration

such as the number of seed-bearing fruits or the number of viable seeds often

decrease with increasing elevation, thereby reducing the probability of seedling

establishment especially above treeline (Cuevas 2000; Kroiss and HilleRisLambers

2015). While seed production and dispersal are unequivocal prerequisites for tree

regeneration, subsequent germination and seedling establishment have also been

widely recognized as potential life-history bottleneck of treeline tree populations

(Stevens and Fox 1991; Germino et al. 2002; Smith et al. 2003; Johnson et al. 2011).

Germination represents the earliest, critical life-stage transition and should thus be

subject to strong natural selection (Baskin and Baskin 2001). Furthermore, the

conditions during germination also influence the phenotypic expression of post-

germination traits, thereby affecting later seedling performance (Donohue et al.


23

2010). Once successfully germinated, the germinant enters the most vulnerable life-

stage of a tree, characterized by the highest mortality of the whole life-cycle (Cui and

Smith 1991; Johnson et al. 2011). Most studies investigating environmental

dependencies of both early life-stages find that favourable conditions are concordant

(i.e. the same conditions are favourable for both, seed and seedling), though others

report conflicting requirements (reviewed in Schupp 1995). Hence, it remains unclear

to what extent the effects of environmental conditions on regeneration success are

life-stage specific.

The natural seedling distribution in treeline ecotones, a result of limitations to

both early life-stages, is often found to be related to stress-reducing site features such

as reduced sky exposure or shelter from strong winds (Germino and Smith 1999;

Smith et al. 2003; Batllori and Camarero 2009). Furthermore, seedling density often

decreases with elevation (Cuevas 2000). Both observations are in line with the view

that the lack of safe sites (sensu Harper 1977) and the harsh climatic conditions in the

alpine zone might restrict the regeneration of treeline trees (Tranquillini 1979). Most

research focusing on the earliest stages of tree regeneration at treeline sites used

germinating seeds principally to study subsequent young seedling survival and

physiology (Germino and Smith 1999; Germino et al. 2002; Moyes et al. 2013), or

lumped germination and subsequent seedling survival due to long observation

intervals (Zurbriggen et al. 2013). Others explicitly including germination responses

at finer temporal scales mainly used elevation gradients to study recruitment

responses, without actively manipulating microclimate (Ferrar et al. 1988; Castanha

et al. 2012). As the process of germination differs in genetic regulation and

environmental sensitivity from survival mechanisms in emerged seedlings and may

Chapter 2

24

thus be evolutionarily decoupled, seedling emergence needs to be monitored

frequently following individual seeds (and seedlings). Moreover, potentially

complex interactions of microclimatic variables and responses of early tree

establishment ask for experimental manipulations of more than one limiting factor.

To our knowledge, no study has ever addressed both of these aspects in a field

experiment on regeneration limitations at treelines.

To summarize, any attempt to understand current treeline patterns and

positions mechanistically and to predict their future dynamics requires investigating

local treelines with regard to microclimate-, species- and life-stage-specific responses.

In this study, within a single field experiment, we assessed the germination and early

seedling-establishment responses of five important European treeline tree species to

the variation of two important microclimatic variables, temperature and moisture.

Accordingly, we asked the following research questions: (a) Do responses of treeline

trees to microclimatic conditions vary with life stage, i.e. during germination and

early seedling establishment? (b) Do temperature and water availability interact to

determine germination and early seedling survival? (c) Do different treeline tree

species show consistent responses to temperature and moisture conditions? In

addition to the field experiment, we monitored the germination response to low

temperatures under controlled conditions in growth chambers for three of the five

study species. This allowed us to assess temperature responses along a defined

gradient and at a finer temporal scale, complementing the results from the more

complex field study.


25

METHODS

Effect of soil moisture and temperature under field conditions

Study site and species

A common garden germination experiment was set up in the experimental garden of

the alpine research station Joseph Fourier in the French Alps near the local treeline

(Lautaret Pass, 2100 m a.s.l., 45°02’N, 6°24’E). The site is situated in a climatic

transition zone between the wet outer Alps and the dry inner Alps (Ozenda 1988),

with 11°C as the mean temperature of the warmest month (July) and an average

annual precipitation of 1230 mm (Choler et al. 2001). The study species comprise four

important treeline-forming conifers of the European Alps: Larix decidua, Picea abies,

Pinus cembra, Pinus uncinata as well as the deciduous angiosperm Sorbus aucuparia,

which also occurs up to treeline elevation (Brändli 1998). Seeds of subalpine origin

from the inner Alps were obtained from a commercial seed producer (Herzog Baum,

Samen und Pflanzen GmbH, Gmunden, Austria) and a forestry office (Kantonaler

Forstgarten Rodels, Rodels, Switzerland), except for seeds of S. aucuparia, which were

available only from colline origin in Hungary (Table 2.1). Information on seed

germinability – either provided by the supplier or determined from standard

germination trials – was used to adjust the seed quantity sown per plot (Table 2.1).

Relatively large seed quantities were sown to account for a potentially lower

germination success under field conditions, allowing a reliable estimation of

germination proportions and ensuring a sufficient number of seedlings to monitor

subsequent survival. Due to time constraints, the seed quantity had to be reduced in

the third experimental block.

Chapter 2

26

Table 2.1. Seed characteristics and seed quantities used in the germination field

experiment

Species Elevation of

seed source

(m a.s.l.)

Germinability (%) Seed quantity (#)

Larix decidua 1800 – 2000 33 % a 240 (120)

Picea abies 1100 – 1400 63 % b 120 (60)

Pinus cembra 1300 - 2850 89 % b 30

Pinus uncinata 2100 78 % a 120 (60)

Sorbus aucuparia 400 – 1400 86 % b 120 (60)

a Germinability of seed lot determined by own standard germination trial (winter 2012/2013)

b Germinability of seed lot provided by seed supplier

Seeds originated always from the inner Alps, except for Sorbus aucuparia, which was only available

from Hungary. Numbers in brackets indicate reduced seed quantity sown in the third experimental

block. For P. cembra seed quantity was always limited to 30 seeds per row due to the large seed size of

1-1.5 cm.

Experimental design

Fifteen experimental plots (70 x 30 cm) were arranged in three blocks to account for

spatial heterogeneity, with approximately 5 m distance between the centers of two

adjacent blocks and 20 cm distance between plot edges. All blocks were enclosed by a

60 cm–high wire-mesh fence as protection against rodents. The vegetation cover on

the plot surface was removed and plots were excavated to a depth of 15 cm to

remove rocks and large roots from the soil. The soil of plots from the same block was

then mixed and returned to the plots. This procedure was done both to create a

homogenous growth substrate within blocks and to remove the effects of biotic

interactions such as competition or facilitation by neighboring vegetation, allowing

to focus on abiotic factors. In October 2013, seeds were sown in one row of 60 cm


27

length per species, allowing 3 cm spacing between rows and 5 cm plot margin. Rows

were randomly assigned to one of the five species. Seeds were sown in 2 cm deep

grooves, distributing seeds evenly with the fingertips and closing up the soil. Seeds

of P. cembra were limited to 30 seeds per row and placed individually due to their

large size.

In spring of 2014, two watering regimes and two types of installations for

passive temperature manipulation, open-top chambers (OTCs; passive warming) and

shade roofs (passive cooling), were set up to create a gradient of soil temperature and

soil moisture across all experimental plots (Fig.2.1A). OTCs were conceived as

hexagons (Marion et al. 1997; r = 80 cm, h = 30 cm) with 3 mm thick acrylic glass

panels transmitting 92 % of solar radiation, including UV. Shade roofs consisted of a

plot-sized wooden frame covered with a shade net, providing 70 % shade on the plot

surface but allowing rain water to pass. The roofs were supported by four 30 cm high

metal poles at the plot corners with 20 cm shade net curtains on each side to prevent

the penetration of low-angle sunshine. Control and warming treatments were

crossed with a watering treatment, with watered plots receiving 3 mm irrigation on

days without rainfall throughout the study period (in total adding up to 35 % of the

May-September precipitation in 2014). Since cooling through shading was already

expected to decrease evapotranspiration and thus increase soil moisture, this

treatment was not included in the additional-watering regime. All five microclimate

treatments (control (C), watered control (C+W), passive warming (OTC), watered

passive warming (OTC+W), passive cooling (Sh)) were replicated in each of the three

experimental blocks. Treatments were initiated directly after snowmelt in a block-

wise manner due to a highly heterogeneous snow cover, with about four weeks

Chapter 2

28

between the start in the first block (mid-April) and the last block (mid-May). The

study period covered the complete growing season of 2014 from snowmelt to early

September.

Fig. 2.1 A. Experimental set-up to manipulate soil temperature and soil moisture by

using shade roofs (on the left, passive cooling) and open-top chambers (on the right,

passive warming) crossed with a watering treatment. B. Individual monitoring of

emerging seedlings by using colored pins (yellow: Larix decidua, green: Pinus cembra,

white: Sorbus aucuparia, red: Pinus uncinata, blue: Picea abies, black: dead seedling).

Microclimate

The soil moisture content (%) was measured monthly with a hand-held sensor

inserted 15 cm in each plot center (TRIME-PICO64, IMKO Micromodultechnik,

Ettlingen, Germany), while the soil temperature was measured at 5 cm depth with

external sensors of permanently installed temperature loggers (Hobo ProV2, Onset

Corp, Bourne, MA, USA). Since comparative measurements in the same

microclimatic treatments in 2013 had shown that there is no significant temperature

difference between watered plots and their control (C vs. W: p = 0.43; OTC vs.

OTC+W: p = 0.76; two-sample t-test, n = 5), soil temperature was only recorded in

the three temperature-relevant treatments C(+W), OTC(+W) and Sh. In each block, a

temperature logger was assigned by chance to either a watered plot or its control and

then inserted in the plot center, recording data in 30 min intervals.

A B


29

Integrated variables were calculated to obtain a quantitative gradient of both

soil moisture and soil temperature across all plots. Mean soil moisture content (soil

MC; %) was calculated as the seasonal average of four monthly measurements for all

plots, giving a soil MC gradient with 15 observation points. Soil heat accumulation

relevant for germination and seedling survival was expressed in growing degree

days with a base temperature (Tb) of 2 °C, which is the lower temperature limit for

germination of at least two of the study species (Løken 1959; Barclay and Crawford

1984). In the plots where temperature was recorded, the number of soil growing

degree days (soil GDD; #) was calculated by summing up the positive differences

between temperature recordings and the base temperature over the whole study

period and dividing the results by the measurement interval fraction of a day (30 min

/ 24 h = 48, Eq. 2.1), giving a soil GDD gradient with nine observation points.

/

( )

. 2.148

i b

i b

i T T

T T

Eq GDD

for temperature recordings (Ti) higher than the base

temperature (Tb).

Seedling survey

Emerging seedlings were recorded weekly and each marked with a colored pin to

allow the assessment of individual survival (Fig. 2.1B). Since hypogeous germination

could not be monitored directly, seedling emergence was used as a proxy for total

germination success by calculating the proportion of sown seeds that emerged as

seedlings (including subsequently dead individuals). Seedling survival was

calculated as the proportion of emerged seedlings that survived until the end of the

growing season.

Chapter 2

30

Effect of temperature under controlled conditions

Complementing the field experiment, a standard germination trial investigating the

effect of low temperatures on germination was performed for three of the five

studied species under controlled conditions in growth chambers (Economic Delux

Snijders Scientific, Thermotec, Weilburg, Germany) in winter 2014 / 2015. Batches of

25 seeds of P. uncinata and P. abies and 60 of L. decidua were placed on moist paper

tissue in sealable plastic boxes (volume = 280 ml) with six replicates for every

temperature treatment. Four low-temperature treatments comprised constant

regimes of 16 °C, 12 °C, 8 °C and 4 °C and all treatments included a 12 h / 12 h light-

dark-cycle. As control treatment we used the settings 20 / 15 °C 12h / 12h,

previously identified as optimal for the same seed lots (germination percentages: P.

uncinata: 79 %, P. abies: 63 %, L. decidua: 33 %). Germination boxes were rotated on

their tray every other day to assure homogeneous temperature exposure.

Successful germination was recorded for every single seed in two- to three-

day intervals as 1 cm growth of the radicle. Germinated seedlings were removed.

The temperature treatments were discontinued when a species showed no further

germination for two weeks. The remaining seeds of the two warmest treatments

(control, 16 °C) were all non-viable, their soft texture and liquid discharge indicating

decay of the embryo, unambiguously indicating maximum germination. Seeds in the

three cooler treatments (12 °C, 8 °C, 4 °C) still showed a very slow increase in

germination after eleven weeks so that remaining, healthy-looking seeds were

transferred to the control temperature for another three weeks to test their viability.


31

Statistical analysis

All analyses were performed using R 3.2.1 (R Core Team 2015). Germination and

survival in the field experiment at the end of the growing season 2014 were

expressed as a two-column vector of counts of successes and failures per species per

plots and analysed with binomial generalized linear models (GLM), including block,

soil MC and soil GDD and the interaction soil MC : soil GDD as explanatory

variables. Using these continuous gradients as explanatory variables instead of the

treatments allowed differentiating the relative effects of soil temperature and

moisture as well as detecting potential interactions between them. In cases of

overdispersion, the standard errors were corrected by using a quasi-GLM model

(Zuur et al. 2009). Non-significant terms were removed from the full models in a

backwards stepwise approach. To facilitate interpretation, significant interactions

between soil MC and GDD are shown graphically by plotting the predicted values

from the model along the whole range of one of these two variables (on the x-axis)

and for three fixed values of the other variable: low (25 % quartile), intermediate

(median) and high (75 % quartile). The variable chosen to represent the x-axis was for

each final model the one with the lower p-value. Note that the resulting curves are

predictions from the models so that they do not directly relate to specific data points;

sections of variables which were not measured are extrapolations of the models.

Since temperature and moisture extremes were also linked to reduced light

intensities by the shading roofs, we evaluated the potentially confounding effect of

light by performing an additional analysis excluding the three shaded plots.

Germination data of the growth chamber experiment were analyzed using

time-to-event analysis (McNair et al. 2012) by first assessing the random variation

Chapter 2

32

among replicates with a Cox proportional-hazards model including a frailty term.

There was no evidence of variability in frailty levels for any of the three species in

any temperature treatment so that data from the six replicates could be pooled. Non-

parametric time-to-event analysis was then used to compare temperature treatment

differences in the germination pattern of each species with a log-rank test using the

survdiff – function in R (survival library; Therneau 2015). Results give test statistics

and significance levels of group (temperature treatment) pairwise comparisons for

each species. The p-values were Holm-adjusted to account for family-wise error rates

in multiple comparisons. Treatment differences of survivor functions were

graphically displayed by showing the inverse Life-table estimates of survivor

functions with point-wise 95 % confidence intervals computed with the R-function

lifetab (KMsurv library; Klein & Moeschberger 2012).

RESULTS

Field experiment: soil moisture and temperature effects on early regeneration

Soil microclimate

The soil moisture content (soil MC) gradient ranged from 20 % to 35.5 %, with the

cooling treatment being the wettest (mean = 33.4 % ± 1.9 SD, n = 3) and the non-

watered control the driest (mean = 25.8 % ± 5.3 SD, n = 3) (Fig. 2.2), with consistent

relative differences between treatments. As expected, heat accumulation of the soil

was highest in the warming treatment (mean = 1313 GDD ± 101 SD, n = 3) and lowest

in the cooling treatment (mean = 769 GDD ± 46 SD, n = 3). The complete gradient

over the nine measured plots ranged from 733 to 1421 GDD (Fig. 2.2).


33

Fig. 2.2 Microclimatic conditions in the field experiment with manipulations of soil

moisture (watering, shading) and soil temperature (passive warming and cooling

treatments). Shown are the integrated microclimate variables mean soil moisture

content (soil MC; %) and total number of soil growing degree days > 2 °C (soil GDD;

#) per treatment (A, C; n = 3) and the respective gradient of soil moisture and soil

temperature over all 15 experimental plots (B, D; sorted by y-axis value, so order

differs for soil MC and soil GDD). Treatment abbreviations indicate: C = Control, W

= watered control, OTC = passive warming (open top chamber), OTC+W = watered

passive warming, Sh = passive cooling (shading roof).

Seedling emergence and survival

Maximum seedling emergence (%) at the end of the growing season under field

conditions was invariably lower than germination under optimum conditions in a

standard germination trial (L. decidua: 25 %, P. abies: 40 %, P. cembra: 63 %, P.

uncinata: 59 %, S. aucuparia: 23 %; see Appendix Fig. S2.1). Overall, seedling survival

at the end of the growing season exceeded an average of 50 % for all species, but

differed considerably among species (L. decidua: 53 %, P. abies: 65 %, P. cembra: 94 %,

P. uncinata: 72 %, S. aucuparia: 67 %; see Appendix Fig. S2.1).

The responses of seedling emergence and first-season survival in the five

study species to soil moisture and soil temperature were highly idiosyncratic. While

So

ilM

C (

%)

So

ilG

DD

(#)

Treatments Experimental plots

A B

DC

Chapter 2

34

higher soil moisture had a positive effect on both stages of early establishment in L.

decidua, it had a generally negative effect on P. cembra (Fig. 2.3, Table 2.2). Similarly,

higher soil temperature generally positively affected seedling emergence in P.

uncinata while having a negative effect on both stages of early establishment in S.

aucuparia (Fig. 2.3, Table 2.2). Within species, however, the principally affecting

climate variable and the direction of its effect were generally consistent for seedling

emergence and survival. There were significant interactions between soil

temperature and moisture in i) the emergence of P. cembra, P. uncinata and S.

aucuparia, as well as ii) the survival of P. abies (Fig. 2.3, Table 2.2): the negative effect

of high soil moisture was reduced (Fig. 2.3B, E) or even inversed (Fig. 2.3D, H) as

temperature increased. Conversely, the negative effects of high temperature were

reduced (Fig. 2.3B, E) or reversed (Fig. 2.3D, H) as the soil moisture content

increased. Finally, a block effect in seedling emergence of S. aucuparia indicated that

emergence was significantly higher in the block with earlier snowmelt (Fig. 2.3E).

An additional analysis excluding the shaded plots showed that effects found

when including all plots were generally maintained even on this shortened

temperature and moisture gradient, with one exception: the significant negative

temperature effects on emergence and survival in S. aucuparia disappeared (see

Appendix Table S2.1).


35

Fig. 2.3 Seedling emergence as proportion of sown seeds (A-E) and survival as proportion of emerged seedlings (F-J) for the five study

species in response to soil moisture (soil MC; %), soil temperature (soil GDD, #) or the interaction of both. Shown are binomial GLM

for significant responses, non-significant responses are displayed as open circles for observed values. Significant interactions are

shown using fixed values of soil GDD (soil MC) plotted along the complete range of soil MC (soil GDD) for P. cembra (P. uncinata, P.

abies, S. aucuparia), line types indicating: dotted = low intensity, dashed = intermediate intensity, solid = high intensity. Note that the

resulting curves are predictions from the models so that they do not directly relate to specific data points. The significant block effect

in seedling emergence of S. aucuparia is displayed by varying hues of grey: black = early snowmelt date (block 1), medium grey =

intermediate snowmelt date (block 2), light grey = late snowmelt date (block 3).

Em

erg

en

ce

Su

rviv

al

Larix decidua Pinus cembra Picea abies Pinus uncinata Sorbus aucuparia

Soil moisture content (%) Soil growing degree days (#)

A B C D E

F

G H I J

Chapter 2

36

Table 2.2 Summary of binomial GLM testing the effect of soil microclimate (soil

moisture content, soil growing degree days) on the proportions of germination and

subsequent survival at the end of the growing season 2014 of five treeline tree species

grown in a field experiment in the French Alps at 2100 m.

Rows give complete models with χ2- or F-values (for GLM and quasi-GLM, respectively) for

germination and survival data at the end of the growing season 2014 with the explanatory variables

block, soil moisture content (Soil MC), soil growing degree days > 2 °C (Soil GDD) and their

interaction (Soil MC : Soil GDD) in the order tested in the model; non-significant variables (given in

parenthesis) were removed from the models based on G- or F-tests, respectively, in a stepwise process

with superscripts indicating the order in which they were removed. The minimum adequate model is

given in bold and significance levels are indicated as: . p<0.1,* p<0.05, ** p<0.01, *** p<0.001. Arrows

indicate whether partial slopes are positive or negative.

Growth chamber experiment: germination response to low temperatures

Germination of the three species was significantly reduced by decreasing

temperatures, but with species-specific differences. In L. decidua the results of the

survivor functions differed significantly mainly between the three highest

temperature treatments, showing a 10 % decrease of germination probability per

treatment (Fig. 2.4A, Table 2.3). Contrasting, in P. abies, the results of the survivor

functions principally differed at the lower end of the temperature gradient (Fig. 2.4B,

Table 2.3), where differences mainly arose from an increasing delay in the onset of

Block Soil MC Soil GDD Soil MC : Soil

GDD

Larix decidua

Germination (F2,9 < 0.01)1 F1,13 = 8.18 * ↑ (F1,12 = 0.31)3 (F1,11 = 0.25)2

Survival (F2,9 = 0.96)1 F1,13 = 7.25 * ↑ (F1,12 = 1.28)3 (F1,11 = 0.27)2

Pinus cembra

Germination (χ2 2 = 4.49)1 χ2

1 = 5.89 * ↓ χ2 1 < 0.01 χ2

1 = 4.91 * ↑

Survival (χ2 2 = 3.01)2 χ2

1 = 10.83 *** ↓ ( χ2 1 = 1.14)3 ( χ2

1 = 0.52)1

Sorbus aucuparia

Germination χ2 2 = 8.1 * χ2

1 = 1.32 χ2 1 = 36.02 *** ↓ χ2

1 = 6.99 ** ↑

Survival (χ2 2 = 1.38)1 (χ2

1 = 0.06)3 χ2 1 = 8.85 ** ↓ ( χ2

1 = 0.36)2

Pinus uncinata

Germination (F2,9 = 0.1)1 F1,11 = 6.84 * ↑ F1,11 = 21.62 *** ↑ F1,11 = 5.11 * ↑

Survival (F2,10 = 0.5)2 (F1,13 = 1.86)4 (F1,1 2= 2.46)3 (F1,9=0.08)1

Picea abies

Germination (F2,12 = 1)4 (F1,10 = 0.53)2 (F1,11 = 1.03)3 (F1,9= 0.08)1

Survival (χ2 2 = 0.85)1 χ2

1= 0.16 χ2 1= 0.55 χ2

1 = 5.75 * ↑


37

germination. Only for the 4 °C treatment the probability of having germinated was

significantly lower (~ 30 % lower) at the end of the experiment than in all other

treatments, though its slope was still positive, potentially indicating a further

increase a longer time span (Fig. 2.4B). In P. uncinata, the most important decrease (~

30 %) in the probability of having germinated occurred at intermediate temperatures,

as was shown by highly significantly different results of survivor functions between

the 12 °C and the 16 °C treatments, while at the high and low end of the gradient,

results were statistically indistinguishable (Fig. 2.4C, Table 2.3).

Viability tests for seeds that did not germinate after eleven weeks in the lower

temperature treatments (12 °C, 8 °C, 4 °C) showed that seed viability was generally

not reduced. In almost all cases, a similar germination success as in the control

treatment was achieved after three additional weeks under control conditions (Fig.

2.4). Only the seeds of L. decidua coming from the 12 °C treatment showed a

substantial (~ 10 %) reduction in their germination success (Fig. 2.4A).

Chapter 2

38

Fig. 2.4 Inverse Life-table estimates of the survivor functions of germination data

from a growth chamber experiment representing the probability of having

germinated at four low temperature treatments and a control (22 / 15 °C, 12h / 12h)

over time, including seeds of L. decidua (A), P. abies (B) and P. uncinata (C). Seeds

were pooled over replicates (n = 6) giving a total of 360 (L. decidua) or 150 (P. abies, P.

uncinata) seeds, respectively. Grey lines represent point-wise 95% confidence

intervals for each treatment. Maximum germination in the control and 16°C-

treatments was achieved after 42 days, with all remaining seeds being non-viable.

After eleven weeks, remaining healthy-looking seeds in the three lower temperature

treatments (12 °C, 8 °C and 4 °C) were transferred to control conditions to test their

viability. Respective treatment symbols at day 100 give mean ± SD of final

germination success after three additional weeks of control conditions.

A Larix decidua

B Picea abies

C Pinus uncinataPro

bab

ilit

y o

fh

avin

gg

erm

ina

ted

Time (days)


39

Table 2.3 Summary of log-rank tests comparing the survivor functions of seeds of

three treeline tree species germinated under controlled conditions in growth

chambers in four permanently low temperature treatments and a control (15 / 22 °C

12h /12h).

Data were pooled over replicates giving a total of 360 (150) seeds for L. decidua (P. abies, P. uncinata),

respectively. Rows show results of group (temperature treatment) pairwise comparisons for each

tested species presenting χ2-values and Holm-adjusted p-values; significant differences are given in

bold, marginally significant differences are given in italics; significance levels are indicated as: .

p<0.1,* p<0.05, ** p<0.01, *** p<0.001.

DISCUSSION

Our results show that the early establishment of the focal treeline tree species is

affected by temperature and water availability in a very idiosyncratic manner.

However, the importance of both climate factors and the direction of their effect on

germination and survival tended to be consistent over both stages of early

establishment within each species. Interactions of both climate variables indicated

that the sensitivity to one factor often depends on the intensity of the other.

Consistency of limiting factors during early establishment

The consistent effect of microclimate over the life-stage transition from germination

to first-year seedling survival (Fig. 2.3) is in accordance with previous studies (Ferrar

Treatment Larix decidua Picea abies Pinus uncinata

comparisons χ2 1 p χ2 1 p χ2 1 p

4 °C vs. 8 °C 0.1 1 61.9 <0.001 *** 7.4 0.17

4 °C vs. 12 °C 15.5 0.08 . 71.8 <0.001 *** 20.8 0.01 *

4 °C vs. 16 °C 0.8 1 89.2 <0.001 *** 125.0 <0.001 ***

4 °C vs. Control 32.1 <0.01 ** 64.3 <0.001 *** 195.0 <0.001 ***

8 °C vs. 12 °C 11.4 0.15 11.3 0.15 5.2 0.3

8 °C vs. 16 °C 0.14 1 28.9 <0.01 ** 100.0 <0.001 ***

8 °C vs. Control 38.8 <0.001 *** 11.2 0.15 177.0 <0.001 ***

12 °C vs. 16 °C 21.0 <0.001 *** 14.0 0.16 71.7 <0.001 ***

12 °C vs. Control 80.5 <0.001 *** 3.6 0.42 147.0 <0.001 ***

16 °C vs. Control 22.4 0.01 * 1.1 1 17.3 0.07 .

Chapter 2

40

et al. 1988; Castanha et al. 2012). This is of particular importance since limitations of a

populations’ distribution are primarily imposed during these most critical life-stages

(Grubb 1977; Harper 1977). In this context, a high level of consistency over two

critical early life-stages will reduce regeneration restrictions arising from seed-

seedling conflicts. On the other hand it should increase the impact of relatively stable

limiting environmental factors, which could be particularly restricting for

regeneration in the harsh conditions of a species’ distribution range edge. In contrast,

a variable factor such as irregular freezing events during the growing season can be

temporarily decoupled from a short, susceptible life-stage such as germination, but is

more likely to affect the longer subsequent stage of the young seedling (Shen et al.

2014). Hence, the degree of concordance or conflict in the environmental

requirements between seed and seedling can have a direct impact on the quantity

and the distribution of recruits (Schupp 1995).

Yet for two species, P. abies and P. uncinata, only one of the two studied life-

stages showed a significant response (Fig. 2.3). This might indicate a change in their

susceptibility to the two microclimatic factors during early establishment, which is

supported by a capability of germinating under a large range of conditions for P.

abies (Løken 1959 and Fig. 2.4B) and a relatively resistant seedling stage in P. uncinata

(Batllori et al. 2010).

Temperature, moisture, and their interactions driving early establishment success

The results from our growth chamber experiment, where decreasing temperatures

invariably reduced germination (Fig. 2.4), are in line with previous studies

demonstrating the importance of temperature for treelines in general (Rolland et al.


41

1998; Körner and Paulsen 2004) and for early regeneration stages in particular

(Germino and Smith 1999; Smith et al. 2003). However, germination success of L.

decidua and P. abies was still considerable at low temperatures (Fig. 2.4A-B). This

concords with the results of our field experiment, where seedling emergence of

neither species was affected by temperature. Inversely, P. uncinata was particularly

temperature-sensitive in the growth chambers and showed a positive temperature

response of seedling emergence in the field, confirming the consistency between both

experiments.

Our field experiment further revealed that almost all seedling emergence and

survival responses were sensitive to water availability, though these responses often

showed an interaction with temperature. In L. decidua moisture was even the only

significant variable, implying that depending on the species, temperature may play

rather a subordinate role in limiting regeneration. Our findings thus add to a

growing body of evidence that other factors than temperature alone, e.g. water

availability, determine seedling distributions at alpine treelines (Ferrar et al. 1988;

Sullivan and Sveinbjörnsson 2010; Greenwood et al. 2015; Kroiss and

HilleRisLambers 2015; Moyes et al. 2015). While germination often requires relatively

high moisture conditions as environmental cue and to set the necessary physiological

processes in motion (Baskin and Baskin 2001), young seedlings depend on it for a

longer period of time due to their generally shallow, simple rooting system and their

large, transpiring surface area relative to the low water storing capacity (Johnson et

al. 2011). Both early life-stages are thus much more affected by water shortage than

established trees, potentially creating a bottleneck for regeneration in treeline and

Chapter 2

42

alpine tundra ecotones, which often exhibit highly variable water holding capacities

(Holtmeier and Broll 2005).

An important result of this study is that the effects of both temperature and

moisture availability on early establishment cannot be decoupled from one another.

Especially the combination of opposite extremes, e.g. high temperatures at low soil

moisture or low temperatures at high soil moisture had limiting effects on both early

life-stages of the study species. Both combinations have previously been shown to

restrict tree development, either by temperature-induced moisture stress (Barber et

al. 2000; Lloyd and Bunn 2007) or cold soil conditions and insufficient aeration

limiting root zone activity (LeBarron 1945; Islam and Macdonald 2004).

Species-specific responses

The observed early-establishment responses to the abiotic environment of the five

tree species were highly idiosyncratic. The temperature response of germination in

the growth chambers revealed a specific pattern for each species (Fig. 2.4, Table 2.3),

possibly indicating an adaption to different ranges of germination temperatures.

These tendencies were confirmed in the field experiment, which further showed

contrasting responses to soil moisture and temperature among all studied species.

Consequently, explaining and understanding observed patterns in the regeneration

limitations of local treeline tree populations requires the consideration of tree species

identity (Wardle 1985; Ball et al. 1991; Sullivan and Sveinbjörnsson 2010; Dufour-

Tremblay et al. 2012) and a detailed connection to the ecology of each individual

species. Hence, in the following we present a tailored, species by species

interpretation of the results.


43

The results of Larix decidua match the ecological features of a typical subalpine,

high-elevation tree species with a high tolerance to cold conditions (see Rameau et al.

1993; Brändli 1998). This was reflected in the relatively high success of germination

down to 4 °C in the growth chambers and the absence of a temperature response for

both early life-stages in the field. Soil moisture, on the other hand, positively affected

seedling emergence and survival (Fig. 2.3A, F), which can be related to the increased

water demand and low water use efficiency of the deciduous life-form compared to

evergreen conifers (Matyssek 1986). This feature seems to be already inherent to the

earliest stages of regeneration, even though first-year seedlings are not deciduous

yet.

In contrast, soil moisture had a negative effect on both early life-stages of

Pinus cembra. This negative effect was for seedling survival but was reduced by

increasing temperatures for seedling emergence. While this appears surprising at

first, it can be explained by a combination of limiting biotic factors and the life-

history strategy of this species. First, P. cembra is the highest-occurring tree species in

Europe and mostly occurs on steep sloping terrain where moisture limitations are

most severe (Brändli 1998). This is not only due to its higher tolerance to the harsh

subalpine conditions, but also caused by its low competitive capacities in relation to

other high-elevation tree species (Ulber et al. 2004). Second, seed dispersal relies on

the nutcracker (Nucifraga caryocatactes), a bird hiding seeds specifically in shallow

caches in open, wind-exposed (Kajimoto et al. 1998) and early snow-free (Mattes

1994) sites. Both of these aspects suggest that P. cembra is well adapted to rather dry

regeneration sites, as corroborated by impressively deep tap roots already present in

small seedlings (Hättenschwiler & Körner 1995, up to 20 cm in 1-yr old plants, i.e.

Chapter 2

44

nearly 4-fold the aboveground plant height, personal observation). Such a rooting

system might, however, be disadvantageous as soil moisture increases since deeper

roots aggravate problems associated with insufficient aeration and cool soil

temperatures (Scott et al. 1987). Furthermore, seedlings are highly vulnerable to

snow fungi promoted by prolonged snow cover (Senn 1999) and in a previous

germination trial (data not shown) we observed high seed mortality due to fouling.

Both findings indicate a general susceptibility of early life-stages to pathogens under

high moisture conditions.

The contrasting responses of early establishment in the other two conifers,

Picea abies and Pinus uncinata, can be directly related to their respective distribution

range. In growth chambers and in the field, the germination response of P. abies was

not or only weakly affected by colder conditions (Figs 2.3C, 2.4B), which is supported

by previous studies reporting germination responses temperatures as low as 2 °C

(Løken 1959). Seedling survival, however, responded to an interaction of

temperature and moisture, with increasing soil moisture compensating a negative

effect of high temperatures (Fig. 2.3H). These findings are in line with the ecological

requirements of this boreal-subalpine tree species, tolerating a wide amplitude of

environmental conditions except drought stress, which is reflected in its absence

from the south side or continental ranges of the European Alps (Rameau et al. 1993;

Brändli 1998).

Pinus uncinata, on the other hand, is a heliophile subalpine tree species with a

southern distribution (Pyrenees, southern European Alps, Rameau et al. 1993) and

accordingly its germination was strongly limited by colder temperature (Fig. 2.4C)


45

and responded positively to higher temperature under sufficient soil moisture

conditions (Fig. 2.3D). Seedling survival was not affected by the manipulated

microclimatic gradients, which is in line with the high tolerance of P. uncinata to

drought and a relatively robust seedling stage (Rameau et al. 1993; Batllori and

Camarero 2009).

Finally, the only broad-leaved and distributionally ubiquitous species Sorbus

aucuparia displayed the counterintuitive response of both early establishment stages

being negatively affected by increasing temperatures. This effect may partly be

explained by a limitation of our study design, in which the plots with shading roofs

(passive cooling) had the coolest temperatures but also an important change in light

conditions. In an additional analysis removing these plots, we showed that the

negative temperature effects on the performance of S. aucuparia disappeared

(Appendix Table S2.1; importantly, removing these plots did not change the general

effects in the conifer species), suggesting that those negative effects were actually an

artefact caused by an increased performance under shaded conditions. This is

supported by the literature, reporting evidence for shade-tolerant seedlings in S.

aucuparia (Raspé et al. 2000; Zywiec and Ledwoń 2008). However, in the case of

seedling emergence, a significant block effect (Table 2.2) indicated a true temperature

effect, i.e. higher emergence with earlier snowmelt, which means colder

temperatures during germination. Furthermore, there is a trend towards higher

seedling emergence in the low-temperature plots even when considering only the

reduced gradient (C and C+W, see Appendix Fig. S2.1). And finally, a true

temperature response for seedling emergence is supported by the germination ability

of S. aucuparia at temperatures as low as 2 °C (Barclay and Crawford 1984) and the

Chapter 2

46

previously found relationship of increasing temperatures reducing germination in an

alpine soil seed bank (Hoyle et al. 2013). Sorbus aucuparia, which possesses traits of

both pioneer and climax species (Zywiec et al. 2013), might benefit from increased

germination at low temperatures in two ways: First, low temperatures could act as an

additional germination cue to increase germination under conditions suitable for

seedling establishment (i.e. shade) and second, it might favour an early germination

time to avoid competing with faster growing species.

Implications for local treeline patterns and dynamics

The regeneration responses found in our study may offer an explanation for

observed patterns and dynamics of treeline tree populations, although such

observations are surprisingly difficult to find in the current literature. For example,

the re-invasion of abandoned subalpine pastures by trees was shown to be restricted

to colluvial soils alongside forest edges for L. dedicua, while being concentrated in

convex relief forms for P. cembra (Didier 2001). According to our results, this may

well be due to the respective early-establishment soil-moisture requirements of these

species (Fig. 2.3A-B, F-G). As another example, water shortage appears not, at least

not yet, to be an issue for recent dynamics of European alpine treelines, since even at

the southern treelines of P. uncinata seedling recruitment is common under current

climatic conditions (Batllori et al. 2010). This species, however, is also known to be

particularly drought resistant (Rameau et al. 1993), while, in line with our results,

boreal spruce forests have already been shown to suffer increasingly from

temperature-induced drought stress (Barber et al. 2000). Accordingly, we found an

important interaction between temperature and soil moisture for seedling survival of

P. abies (Fig. 2.3H). Hence, depending on local changes in precipitation, growth and


47

recruitment of high-elevation populations of this species could become restricted by

a warming climate even though they were, until recently, positively affected by it

(Bolli et al. 2007) and thereby affect the responsiveness of species-specific treelines to

increasing temperatures. However, note that a direct relation to local treeline features

will remain difficult, because many treelines are subject to intense anthropogenic

influences and are currently not at their climatic limit (Wick and Tinner 1997). Land

use can thus be a primary driver of their spatial pattern and recent dynamics, in

particular in the European Alps (Didier 2001; Bolli et al. 2007; Vittoz et al. 2008).

Therefore, in addition to climatic factors, land-use history needs to be taken into

account in observational studies of treeline dynamics.

Our results can also be linked to the important contribution that species-

specific requirements of the earliest life-stages exert on the shape and dynamic of a

local treeline (Harsch and Bader 2011). For example, if young seedlings require shade

or shelter – as did S. aucuparia in our study – they will be most successful near

existing trees and treeline tree populations will tend to occur in clustered spatial

patterns (Smith et al. 2003). Conversely, species requiring increased temperature or

light conditions – such as P. cembra and P. uncinata according to our results (Fig. 2.3)

– may perform better in open microsites and their treeline populations may develop

a scattered distribution (Holtmeier 2009). Consequently, abrupt treelines, if not

caused by disturbances, are primarily explained by high seedling mortality beyond

the forest edge and less so by growing season temperatures, which makes them less

responsive to the current climate change. In diffuse treelines in contrast, growth is

more and more limited by temperature with increasing elevation or latitude, and

Chapter 2

48

accordingly most treeline advances can be expected in this treeline type (Harsch and

Bader 2011).

CONCLUSIONS

Recruitment as a population bottleneck plays a crucial role in the discussion about

driving forces and future dynamics of treeline ecotones. To our knowledge this is the

first study to link the two earliest stages of tree establishment in a multi-species

approach experimentally manipulating two potentially limiting microclimatic

variables. We show that responses are highly idiosyncratic, but generally consistent

over both life-stages within each species, which increases the impact of limiting

climate variables in a relatively stable environment. Furthermore, interactions of

temperature and moisture highlight the complex interplay of microclimatic factors

influencing the regeneration success and confirm the importance of other factors than

temperature, such as water availability, for the understanding of treeline dynamics.

Our study contributes to the understanding of species-specific requirements and

limitations of the vulnerable stages of early establishment, which can be used to

explain current treeline patterns and predict future responses in the context of their

local climatic conditions.

ACKNOWLEDGEMENTS

We are very grateful to Serge Aubert, former director of the Lautaret Alpine

Botanical Garden and the affiliated research station Joseph Fourier, for enabling this

study by allowing us to use the experimental space and facilities of the station, and

thank the gardener team for support with any technical problem. We also thank


49

several field assistants, especially Carla Sardemann and Mathilde Vicente, for their

help in installing, maintaining and monitoring the experiment. The study was

funded by the German Research Foundation (DFG, BA 3843/5-1&2).

Chapter 2

50

APPENDIX

Fig. S2.1 Proportions of seedling emergence and subsequent survival (mean ± SD; n = 3) at the end of the first growing season for five

treeline tree species grown in a field experiment with manipulations of soil moisture and soil temperature in the French Alps at 2100

m a.s.l.. Treatment abbreviations indicate: C = Control, W = watered control, OTC = passive warming (open top chamber), OTC+W =

watered passive warming, Sh = passive cooling (shading roof). Different letters stand for, where present, significant differences. Note

that there are no error bars for emergence of Pinus cembra, OTC (C), because the number of seedlings was the same in all three

replicates.

Em

erg

en

ce

Su

rviv

al

Treatments

Larix decidua Pinus cembraPicea abies Pinus uncinata Sorbus aucuparia

A B C D E

F G

H

I J

a

b

abab

ab

aa

b

ab

ab

a

bb

abab

bb b

ab a


51

Table S2.1 Summary of binomial GLM testing the effect of soil microclimate (soil

moisture content, soil growing degree days) for reduced gradients (excluding plots

with shading roofs) on the proportions of seedling emergence and subsequent

survival at the end of the growing season 2014 of five treeline tree species grown in a

field experiment in the French Alps at 2100 m.

Rows give complete models with χ2- or F-values (for GLM and quasi-GLM, respectively) for

germination and survival data at the end of the growing season 2014 with the explanatory variables

block, soil moisture content (Soil MC), soil growing degree days (> 2 °C, Soil GDD) and their

interaction (Soil MC : Soil GDD) in the order tested in the model as indicated by superscripts; non-

significant variables (given in parenthesis) were removed from the models based on G- or F-tests,

respectively, in a stepwise process with superscripts indicating the order in which they were removed.

The minimum adequate model is given in bold and significance levels are indicated as: . p<0.1,*

p<0.05, ** p<0.01, *** p<0.001. Arrows indicate whether partial slopes are positive or negative.

Block Soil MC GDD Soil MC : GDD

Larix decidua

Germination (F2,6 = 0.01)1 F1,10 = 11.2 * ↑ (F1,9 = 0.26)3 (F1,8 = 0.42)2

Survival (F2,6 = 0.32)1 F1,10 = 9.46 * ↑ (F1,9 < 0.01)3 (F1,8 = 0.13)2

Pinus cembra

Germination χ2 2 = 6.75 * (χ2 1 = 2.65)3 (χ2 1 = 0.47)2 (χ2 1 = 2.66)1

Survival (χ2 2 = 0.42)2 χ2 1 = 6.47 * ↓ ( χ2 1 = 1.95)3 ( χ2 1 = 0.02)1

Sorbus aucuparia

Germination (F2,6 = 0.99)1 (F1,9 = 0.67)3 (F1,10 = 1.19)4 (F1,8 = 0.96)2

Survival (χ2 2 = 0.95)3 ( χ2 1 = 0.12)2 (χ2 1 = 0.23)4 ( χ2 1 = 0.1)1

Pinus uncinata

Germination (F2,6 = 0.58)1 F1,9 = 4.36 . ↑ F1,9 = 7.32 * ↑ (F1,9= 1.11)2

Survival (F2,7 = 1.29)2 F1,9 = 7.39 * ↑ F1,9 = 8.28 * ↑ (F1,6=0.03)1

Picea abies

Germination (F2,9 = 1.43)4 (F1,7 = 0.01)2 (F1,8 = 3.23)3 (F1,6 = 0.06)1

Survival (χ2 2 = 4.45)4 (χ2 1= 0)2 (χ2 1= 0.5)3 (χ2 1= 0.26)1

53

CHAPTER 3

COMPETITOR OR FACILITATOR? THE ROLE OF GRASSLAND

VEGETATION FOR GERMINATION AND SEEDLING PERFORMANCE OF

TREE SPECIES AT THE ALPINE TREELINE

Hannah Loranger, Gerhard Zotz, Maaike Y. Bader

Submitted

ABSTRACT

Alpine treelines constitute conspicuous vegetation boundaries in mountain

ecosystems that are expected to move upslope with a warming climate. However,

treeline responses are inconsistent and remain difficult to predict since many factors

unrelated to temperature, such as biotic interactions, can influence them on a local

scale. Especially during early regeneration stages, tree seedlings can be strongly

Chapter 3

54

influenced by alpine herbaceous vegetation through both competition and

facilitation. We aimed to understand the relative importance of these two types of

interactions, in dependence of vegetation structure, for treeline regeneration

dynamics. We studied the effect of herbaceous alpine vegetation on germination and

first-year seedling performance in a field experiment in the French Alps (2100 m

a.s.l.) with five important European treeline tree species: Larix decidua, Picea abies,

Pinus cembra, Pinus uncinata and Sorbus aucuparia. We focussed on how reserve

storage and allocation, studied via measurements of non-structural carbohydrates

and seedling biomass, are affected by varying vegetation cover and how this

interaction changes seasonally during the first year. The results show the dominance

of negative vegetation impacts, including general competition effects as well as tree-

species-specific susceptibilities to combinations of competition and indirect

vegetation effects via microclimate or pathogens. However, evergreen tree seedlings

appear to benefit from protection by the senescent herbs in autumn, leading to

increased carbohydrate reserves at the end of the winter. Thus, the interaction with

herbaceous vegetation switches seasonally from competition to facilitation. It is

currently unclear whether this effect promotes long-term net facilitation for tree

seedlings or if competitive interactions with the herbaceous vegetation prevail,

highlighting the need of long-term studies evaluating the impact of biotic

interactions at alpine treelines. Since early regeneration determines whether treelines

remain stable or move upslope, our findings contribute to the understanding of

observed treeline patterns, dynamics and their local variation in the context of site-,

life-stage and species-specific processes. This has important implications for the

development of predictive models of treeline dynamics, in which these ‘local’ aspects

Effects of Herbaceous Vegetation

55

should be incorporated in addition to more global drivers like changes in

temperature.

Keywords: biotic interactions, early establishment, Larix decidua, non-structural

carbohydrates, Picea abies, Pinus uncinata, Pinus cembra, Sorbus aucuparia

Chapter 3

56

INTRODUCTION

The fast, ongoing changes in global climate have already caused important shifts in

species distributions, which are predicted to continue in the future (Parmesan and

Yohe 2003; Chen et al. 2011). Treeline ecotones form the most conspicuous vegetation

boundary in alpine and arctic ecosystems and are particularly interesting in the light

of distributional shifts, since an entire life-form – the tree – reaches a thermal limit

(Tranquillini 1979; Körner 1998). This global pattern led to the prediction of an

upslope, or in arctic systems poleward, migration of tree cover with a warming

climate (Grace 2002). However, observations of local treeline positions and recent

dynamics show contrasting patterns (Holtmeier and Broll 2005; Harsch et al. 2009;

Case and Duncan 2014). While climatic factors are important in driving species

responses to climate changes, they interact with non-climatic factors in a close and

complex manner (Sutherst et al. 2007; Brown and Vellend 2014). Disregarding non-

climatic factors when considering distributional shifts, in particular species-specific

responses and species interactions, can therefore lead to unreliable predictions (Davis

et al. 1998).

Biotic interactions belong to such non-climatic factors with the potential to

limit or extend a species’ ecological and geographic range (Wisz et al. 2013 and

references therein). They can thus be an important factor causing variation in local

treeline responses, as shown by the impact of soil biota (Hasselquist et al. 2005),

animals (via herbivory and disturbances, Cairns and Moen 2004) and neighbouring

plants (e.g. Hobbie et al. 1999; Akhalkatsi et al. 2006; Grau et al. 2012) on the

distribution of trees at their elevational or latitudinal range edge. Plant-plant


57

interactions might be particularly important for the successful establishment of

seedlings, with neighbouring vegetation either causing competition (Venn et al. 2009)

or facilitating seedling growth (Sullivan and Sveinbjörnsson 2010). Under stressful

conditions the importance of positive (facilitative) interactions tends to increase

relative to negative (competitive) interactions (the ‘stress gradient hypothesis’, SGH,

Bertness and Callaway 1994). While several studies provide support for the SGH in

alpine plant communities (Choler et al. 2001; Callaway et al. 2002; He et al. 2013 and

references therein), it has recently been argued that species-interactions can switch

back to competition at the extreme ends of a stress gradient (Michalet et al. 2014). For

the alpine treeline, the interaction between existing alpine vegetation and

establishing tree seedlings is likely a balance between positive and negative effects,

depending on the alpine vegetation type, the local climatic conditions, and the

tolerance limits of the tree species involved.

Herbaceous vegetation may facilitate tree germination and seedling

performance at treeline in many ways, e.g. via increased canopy and soil

temperatures through passive solar warming and soil insulation (Körner and Paulsen

2004), or via decreased exposure to high solar radiation through shading (Germino et

al. 2002; Maher et al. 2005; Bader et al. 2007), reducing the risks of desiccation and

photoinhibition (Ball et al. 1991; Germino and Smith 1999). Furthermore, it can

reduce desiccation stress and mechanical damage produced by wind and by frost

heave, with increased seedling survival compared to open patches (Noble and

Alexander 1977; Carlsson and Callaghan 1991; Ryser 1993). On the other hand, for

seedlings reaching above the herbaceous canopy or growing in open patches

surrounded by vegetation, temperature extremes may be exacerbated compared to

Chapter 3

58

bare ground due to radiative warming and cooling of the vegetation (Ball et al. 1997;

Germino et al. 2002). Neighbouring grasses and herbs can also be strong competitors

for resources such as light, water and nutrients (Nambiar 1990; Moir et al. 1999).

These contrasting effects of herbaceous vegetation appear to strongly depend on

vegetation height and density.

Many of the potential effects of alpine vegetation on seedling performance are

related to the seedling’s carbon balance. Adult trees at treeline do not appear to be

limited by carbon (e.g. Hoch and Körner 2003; Shi et al. 2008; Fajardo et al. 2012;

Molina-Montenegro et al. 2012). Seedlings, however, differ substantially from adult

trees in terms of physiology (Day et al. 2001; Tegischer et al. 2002) and the

microclimate they experience (Körner 1998). More specifically, treeline trees in their

early establishment may be subjected to greater carbon limitations, since they have

only small amounts of productive and storage tissues, resulting in a risk of carbon

starvation (Wang and Zwiazek 1999; Li et al. 2002). Furthermore, seedlings are

covered under snow well into spring at most non-tropical treelines, resulting in a

much shorter growing season for reserve accumulation and tissue maturation

(Bansal and Germino 2009). Finally, young seedlings may be more susceptible to

cold-induced photoinhibition (Ball et al. 1991; Germino and Smith 1999; Bader et al.

2007), limiting photosynthetic efficiency and lowering photosynthesis when

sustained (Close et al. 2000). All these limitations can be mitigated or aggravated by

the potentially positive or negative impacts that alpine vegetation exerts on young

tree seedlings’ resource acquisition (Germino et al. 2002; Maher and Germino 2006).


59

The aim of our experimental study was to assess the balance of positive and

negative effects of alpine herbaceous vegetation on early seedling establishment in

different treeline tree species. Our research questions were: (a) How do varying

intensities of herbaceous vegetation cover influence the microclimate experienced by

treeline tree seedlings? (b) How do seedlings of five common treeline tree species

respond to varying vegetation cover? c) How is the carbon balance during early

establishment affected by adjacent herbaceous vegetation, and how do these effects

differ between seasons and species?

MATERIALS AND METHODS

Study site and species

The study site was located in the experimental garden of the Alpine Research Station

Joseph Fourier in the French Alps near the local treeline (Lautaret Pass, 2100m a.s.l.,

45°02’N, 6°24’E). The site is situated in a climatic transition zone of the wet outer

Alps to the dry inner Alps (Ozenda 1988), with 11°C as the mean temperature of the

warmest month (July) and an average annual precipitation of 1230 mm (Choler et al.

2001). In winter permanent snow cover typically lasts 4-5 months with a moderate

depth of 2-3 m (Franck Delbart, Alpine Research Station Joseph Fourier, personal

communication). Due to past transformation of natural forests to pastures there is no

natural treeline at the Lautaret pass and the vegetation is dominated by species-rich

Festuca paniculata – meadows (Quétier et al. 2007).

The study species comprise the four important treeline-forming conifers Larix

decidua L., Picea abies (L.) Karst, Pinus cembra L., Pinus uncinata Ram. as well as the

broadleaved Sorbus aucuparia L., which also occurs up to treeline elevation in the

Chapter 3

60

Alps (Brändli 1998, personal observation). Seeds of subalpine origin were obtained

from a commercial seed producer (Herzog Baum, Samen und Pflanzen GmbH,

Gmunden, Austria) and a forestry office (Kantonaler Forstgarten Rodels, Rodels,

Switzerland) providing seeds from the inner Alps, except for seeds of Sorbus

aucuparia, which were available only from colline origin in Hungary. Seeds were

germinated beforehand in plastic trays with a homogeneous mixture of commercial

propagation substrate and sand under cool greenhouse conditions (10 – 15 °C) in

March 2013 in Oldenburg, Germany. After approximately one month, seedlings were

transported to the study site and kept indoors at the same temperature regime as in

the nursery due to harsh winter weather until the end of May 2013. Prior to

transplanting into the experiment in early June 2013, seedlings were subjected to an

acclimatization period of three weeks in lightly shaded nursery beds (roofs with

shading cloths providing 40 % shade and allowing penetration of direct sunlight in

the evening) of the Lautaret Alpine Botanical Garden.

Experimental design

Twelve experimental plots (70 x 30 cm) were arranged in three blocks to account for

spatial heterogeneity, with four treatments per block. There was approximately 5 m

distance between the centers of two adjacent blocks and 20 cm distance between

plots. All blocks were surrounded by a 60 cm high wire-mesh fence as protection

against rodents. The treatments consisted of three levels of vegetation cover (Fig. 3.1):

full vegetation cover (FV), intermediate vegetation cover (IV) and bare ground (BG).

Additionally, a shaded bare ground treatment (Sh) was included to study the effect

of light reduction without further biotic interactions between tree seedlings and

herbaceous vegetation. In the full vegetation treatment, the locally occurring dense,


61

intact alpine grassland vegetation was left unmanipulated. These plots had a

vegetation height at peak biomass in mid-July of approximately 25 cm and a PAR

(photosynthetic active radiation) reduction of 80 % compared to bare ground

(measured at seedling height in the plot centres with a PAR-sensor, Microstation,

Onset Corp, Bourne, MA, USA). For the intermediate vegetation treatment, the

herbaceous vegetation was reduced by regular cutting to a height of 7 cm, which was

approximately seedling height. In this treatment, tree seedlings were largely relieved

of light competition (PAR reduction of 10 % compared to BG) but still subject to

belowground competition and a microclimate essentially modified by surrounding

vegetation. For the bare ground treatments with and without shading (BG and Sh),

plots were excavated to a depth of 15 cm to remove all above- and belowground

plant material. The soil of both vegetation-free plots of a block was then mixed and

returned, creating a homogenous growth substrate within blocks. The shaded

treatments were covered with shading roofs, i.e. plot-sized wooden frames covered

with shade cloth creating light conditions similar to the full vegetation treatment (70

% PAR reduction compared to BG) while allowing rain water to pass. The roofs were

supported by four 30 cm high metal poles at the plot corners and featured 20 cm

shade-cloth curtains on each side to prevent the penetration of low-angle sunshine.

Tree seedlings were planted into the treatment plots in 60-cm rows, with one row per

species and ten seedlings per row. This allowed for 6 cm spacing within and 5 cm

between rows. After planting, seedlings were watered daily for two weeks to reduce

planting stress and associated mortality. Seedling survival was recorded monthly

throughout the growing season and once again in the following spring after

snowmelt.

Chapter 3

62

Fig. 3.1 Experimental set-up including three levels of vegetation cover: A. Full

vegetation cover (no manipulation of the initial alpine grassland vegetation after

seeding /planting of the seedlings). B. Intermediate vegetation cover (reduction of

the vegetation by regular cutting to a height of approximately 7 cm). C. Bare ground

(removal of all above- and belowground plant material). Note the protective shield of

the humidity sensor (10 cm height, furthest left white circle), which is freely visible in

the intermediate vegetation treatment and almost completely overgrown in the full

vegetation treatment (lower left side). Pictures were taken at peak biomass in mid-

July 2014.

This seedling experiment was accompanied by a seed germination

experiment. Seeds of the same five tree species, originating from the same seed

sources, were sown in the four treatments (FV, IV, BG and Sh) in autumn 2013 as

described in Loranger et al. (2016, submitted). Starting after snowmelt in spring 2014,

emerging seedlings were recorded weekly and marked with a colored pin to allow

monitoring each individual. Since germination of soil-covered seeds could not be

monitored directly, seedling emergence was used as a proxy. Germination rates were

defined as the total proportion of emerged seedlings from sown seeds (including

subsequently dead individuals, at the end of the growing season in September 2014).

Microclimate

The microclimate within treatments was monitored by measuring air and soil

temperature with permanently installed temperature loggers (Hobo ProV2, Onset

Corp, Bourne, MA, USA) featuring two external sensors at 5 cm belowground and,

A B C


63

protected by a sunshield, at 10 cm aboveground, over the complete growing season

from June to September. Relative air humidity (%) was measured using loggers

(HumiLog “rugged”, Driesen & Kern GmbH, Bad Bramstedt, Germany) with an

external humidity sensor protected by a rain/sun shield. All sensors were placed in a

central position in the plot and data were recorded in 30-minute intervals. Due to

limited numbers of HumiLog loggers, air humidity was measured only in three

treatments out of four – i.e. FV, IV, and BG – and in two treatments simultaneously,

comparing IV and BG treatments, respectively, with FV in two subsequent

measurement periods of six days. Furthermore, soil moisture content (%) was

measured monthly throughout the growing season with a hand-held sensor inserted

15 cm deep in each plot center (TRIME-PICO64, IMKO Micromodultechnik,

Ettlingen, Germany).

Air and soil temperatures were recorded in 2013 and 2014, but due to logistical

problems air humidity and soil-moisture contents were recorded only in 2014 and

were therefore used only to determine relative differences between treatments. Air

and soil temperatures showed the same treatment effects for 2013 and 2014. Annual

cumulative precipitation measured by the nearest weather station were very similar

in both years, with 970 mm and 932 mm, respectively (Besse en Oisan Weather

Station Météo France, 1525 m a.s.l., 45°04’N, 6°10’E). Microclimate data are therefore

presented for 2014 only.

Biomass sampling and carbohydrate analysis

Complete seedlings were harvested to determine seedling biomass and carbohydrate

concentrations. Sampling took place twice, at the end of the first growing season

Chapter 3

64

(September 11 – 13, 2013, one block per day) and in the following spring three-to-six

days after snowmelt (May 19 – 22, 2014). For the deciduous species S. aucuparia this

timing was before autumn senescence of leaves and after the opening of leaf buds in

spring. The evergreen conifer species had only first-year needles in both sampling

periods, since their leaf buds were still closed during the spring sampling. This

includes also the generally deciduous conifer L. decidua, whose seedlings do not shed

their needles during the first years. In each sampling, one seedling per species per

plot was randomly selected, carefully excavated and immediately stored on ice.

Samples were brought to the field laboratory within 2 hours, where roots were

washed thoroughly and samples were then microwaved for 90 s at 600 W to stop

enzymatic reactions (Popp et al. 1996). Samples were then dried to constant weight at

70 °C for 48 h and stored on silica gel in sealed containers until further analysis.

Additionally, the initial seedling biomass before planting was determined for five

randomly selected seedlings of each species (L. decidua: 14.8 ± 3.9 mg, P. abies: 22.5 ±

4.6 mg, P. cembra: 170.6 ± 47.7 mg, P. uncinata: 11.6 ± 3.2 mg, S. aucuparia: 24.4 ± 5.5

mg, mean ± SD, n = 5).

Non-structural carbohydrates (NSC) are defined here as the sum of soluble,

low-molecular-weight sugars (glucose, sucrose, fructose and maltose) and starch. The

chemical analysis of NSC took place in the Functional Ecology laboratory of the

University of Oldenburg, where biomass samples were separated into leaf, stem and

root biomass and weighed to ± 0.01 mg. The separated seedling part samples were

than ground to a fine powder and soluble sugars were extracted by heating

approximately 5 mg of this powder mixed with distilled water for 30 min at 80 °C.

After three consecutive centrifugation and dilution steps to wash out soluble sugars


65

from the sample residue, the combined supernatant was used for measuring soluble

sugar concentration via High Performance Liquid Chromatography (HPLC, ICS-

3000, Dionex Corporation, Sunnyvale, California). For starch, each residue from the

first extraction was mixed with α-amylase, amyloglucosidase, distilled water and

acetate buffer and heated for 2 h at 30 °C to break down starch into glucose. Again,

after three consecutive centrifugation and dilution steps, glucose concentration

originating from starch was measured using HPLC. Concentrations (mg g-1 dry

weight) of each sugar were determined via comparison with standards. It is

important to note that results of NSC-analyses were shown to differ importantly

between laboratories and applied methods (Quentin et al. 2015), so that no cross-

comparisons of absolute values are possible without cross-calibration. Results

originating from the same laboratory and the same method, however, should be

comparable.

Statistical analysis

All analyses were performed using R 3.2.2 (R Core Team 2015). Treatment effects on

plot microclimate were evaluated by analysing seasonal averages of daily minimum,

maximum and mean air and soil temperatures as well as seasonal averages of soil

moisture content with single-factor analysis of variance (ANOVA). For relative air

humidity, which was obtained for only two treatments simultaneously, replicates

were in time (i.e. daily minimum, maximum and mean per day of measurement) and

differences were analysed with paired t-tests for each experimental block separately.

Survival of transplanted seedlings over summer (September 2013) and winter

(May 2014), and total seasonal germination of sown seeds (September 2014) were

Chapter 3

66

expressed as two-column-vectors of successes and failures for each species and

analysed with binomial generalized linear models (GLM), including block and

treatment as explanatory variables. In cases of overdispersion, the standard errors

were corrected by using a quasi-GLM model (Zuur et al. 2009).

At the level of whole seedlings, differences in biomass and carbohydrate

reserves (total NSC concentrations) between treatments and seasons were tested

using two-way factorial ANOVA or analysis of covariance (ANCOVA) for the latter,

controlling for size-effects by including seedling biomass as covariable. On the level

of separated seedling parts, differences in total soluble sugars, starch and NSC were

analysed with an ANCOVA including treatment, part and season as independent

factors with their respective two-way interactions, and biomass as covariable. Where

necessary, data were transformed to meet model assumptions and non-significant

terms were removed from the full models in a backwards stepwise approach. All

statistically significant results (p < 0.05) were evaluated using Tukey honestly

significant difference criteria for pairwise comparisons with Bonferroni corrections

for multiple comparisons. For each species, response variables were analysed in a

separate model, and block was included as a factor in all models to account for the

spatial structure of the experiment.

RESULTS

Microclimate

The four treatments applied in this study (BG, IV, FV, Sh) significantly affected the

microclimate experienced by the germinating and transplanted seedlings. The soil

moisture content in the bare ground treatment was consistently and significantly


67

lower than in the other treatments (26 ± 5 % vs. 33 ± 3 %; means ± SD; F3,30 = 5.74, p <

0.01). Relative air humidity (rh) was generally lower in the bare ground and the

intermediate vegetation compared to the full vegetation treatment (daily minimum

rh in BG vs. FV: 45 ± 2 vs. 61 ± 8 % and IV vs. FV: 46 ± 3 vs. 56 ± 3 %, means ± SD, n =

3; see Appendix Table S3.1).

Mean and maximum daily air temperatures were significantly lower in the

shading and full vegetation treatment and higher in the intermediate vegetation

treatment compared to bare ground (Fig. 3.2B-C). The minimum daily air

temperature, however, varied much less; the only significant effect being that

shading was colder than full vegetation (Fig. 3.2A). Overall the intermediate

vegetation treatment had the largest air temperature fluctuations (Fig. 3.2). The mean

daily soil temperature was lowest in the shading treatment and highest in the bare

ground and intermediate vegetation treatments (Fig. 3.2C). Thus, soils under

artificial shading were generally coldest, while the bare ground treatment

experienced both the warmest and the second-coldest temperatures and had the

largest soil temperature fluctuations. The soil under the intermediate vegetation

treatment was significantly warmer than the full vegetation treatment (Fig. 3.2), but

both treatments provided a buffering effect leading to relatively low maximum and

high minimum soil temperatures, i.e. small daily soil-temperature fluctuations (Fig.

3.2A-B).

Chapter 3

68

Fig. 3.2 Seasonal means (± SD, n = 3 plots) of daily minimum (A), maximum (B) and

mean (C) temperatures measured in four treatments with manipulations of

vegetation cover and microclimate in a seedling transplant field experiment at 2100

m a.s.l. in the French Alps over the whole growing season (June 6 to September 4,

2014). Open circles indicate air, filled circles soil temperatures. Different letters

indicate significant differences, with upper case letters for air temperatures and

lower case italic letters for soil temperatures. BG = Bare ground (no vegetation

cover), IV = Intermediate vegetation cover, FV = Full vegetation cover, Sh = Shading

(no vegetation cover).

Seedling performance and total reserve accumulation

The only treatment significantly affecting seedling survival was the full vegetation

cover, decreasing survival in three of the five species (L. decidua: F3,19 = 5.3, p<0.01; P.

uncinata: F3,18 = 24.3, p<0.001; S. aucuparia: F3,18 = 6.9, p<0.01). While in both conifers

all dead seedlings were at least partly covered by mould fungi, suggesting a causal

relationship, there was no visible cause of death in S. aucuparia. Moreover, survival

was significantly reduced in L. decidua and P. abies after the winter in all treatments

(F1,19 = 5.2, p<0.05) (F1,20 = 13.5, p<0.01). Pinus cembra was not affected by any

treatment and there was no additional winter mortality (Fig. 3.3).

Treatments

Tem

pe

ratu

re(°

C)

B

a

bc

d

AB AB A

A

a ab

c

AB

C C

CB

a

b

c c

A

B

CC


69

Fig. 3.3 Means (± SD, n = 3) of survival percentage (first row), total seedling biomass

(second row) and total non-structural carbohydrates (NSC) concentrations per

seedling (third row). Seedlings were grown in a field experiment with manipulations

of vegetation cover and microclimate in the French Alps at 2100 m a.s.l. and samples

were taken in two subsequent seasons: at the end of the first summer (September

2013, black bars) and at the end of the first winter (May 2014, grey bars). Seedling

biomass did not differ between seasons and the data were therefore pooled over

seasons (no extra growth over the winter). Different letters stand for significant

treatment differences, significant seasonal differences or interactions between

treatments and season are represented by the symbols S* and I*, respectively. BG =

Bare ground (no vegetation cover), IV = Intermediate vegetation cover, FV = Full

vegetation cover, Sh = Shading (no vegetation cover).

Seedling biomass was lower in full and intermediate vegetation compared to

bare ground and shading for L. decidua, P. uncinata, S. aucuparia and P. abies, whereas

in the latter the effect of intermediate vegetation was less pronounced (Table 3.1, Fig.

3.3). Pinus cembra was again exceptional: all vegetation- or roof-covered treatments

resulted in lower seedling biomass compared to the bare ground (Table 3.1, Fig. 3.3).

Bio

mas

s(m

g)N

SC c

on

cen

trat

ion

s(m

g g-1

dw

)Su

rviv

al(%

)

Treatments

S*

aa

a

b

a

a

a

b

I*

aa

b

b

a

c

ab

bc

a

b bb

aa

b

b

a

c

ab

bc

abab

aba a a

b

S*Larix decidua Picea abies Pinus cembra Pinus uncinata Sorbus aucuparia

a a ab

S*

BG BG BG BG BG

Chapter 3

70

Table 3.1. Summary of ANOVA/ANCOVA results testing the effect of vegetation

cover (comprising bare ground, bare ground with shading, intermediate and full

vegetation cover as treatments) on the total biomass and total NSC concentrations of

seedlings of five treeline tree species grown in a field experiment in the French Alps

at 2100 m. The two sampling seasons were the end of the first growing season and

the end of the first winter in 2013/2014.

Total Biomass Total NSC

F P F p

Larix decidua

Treatment 9.05 <0.001 9.04 <0.001

Season 0.37 0.55 1.30 0.27

Treatment:Season 0.63 0.61 1.65 0.22

Picea abies

Treatment 8.73 <0.001 0.34 0.80

Season 1.03 0.32 0.01 0.90


Pinus cembra

Treatment 10.22 <0.001 0.33 0.80

Season <0.01 0.95 0.15 0.70

Treatment:Season 0.46 0.71 6.02 <0.01

Pinus uncinata

Treatment 15.51 <0.001 7.83 0.001

Season 5.11 0.04 0.49 0.49


Sorbus aucuparia

Treatment 8.59 <0.001 0.89 0.47

Season 0.52 0.48 7.14 0.01


Shown are the main effects treatment and season as well as their interaction for both response

variables total biomass and total NSC concentrations of each species reporting F- and p-values.

Significant effects are given in bold and italics.

The total non-structural carbohydrate (NSC) concentration per seedling was

lower in full vegetation compared to all other treatments in L. decidua and P. uncinata.

In P. cembra, an interaction between treatment and season indicated that this was the

case only at the end of the summer, while seedling NSC concentration in full

vegetation increased over the winter half-year in contrast to the other treatments


71

(Table 3.1, Fig. 3.3). A similar pattern was found in the other three conifers (Fig. 3.3).

In the only broadleaved species, S. aucuparia, total NSC concentrations were not

affected by treatments, but decreased significantly over the winter half-year, while P.

abies was neither significantly affected by treatment nor by season (Table 3.1, Fig.

3.3).

None of the different treatments affected seedling emergence, except for a

positive effect of shade roofs on S. aucuparia (F3,8 = 4.6, p<0.05, data not shown).

Organ-dependent reserve accumulation

The patterns of reserve accumulation in seedling parts as a function of treatment and

season is consistent with that of entire seedlings. For the four conifers, the main

effects indicated that i) seedlings in the full vegetation treatment had generally lower

carbohydrate concentrations compared to the other treatments, ii) reserve

accumulation (especially starch) was highest in the roots, and iii) starch reserves

decreased over winter (Fig. 3.4, Appendix Table S3.2). Carbohydrate concentrations

within treatments and seedling parts often differed between seasons, but patterns

were plant-part and species specific, which is shown by many highly significant

interactions (Fig. 3.4, Appendix Table S3.2).

In the following paragraphs we present more specific results. Frequently,

trends are similar, but not always significant for all NSC components and species.

Hence, for conciseness, the NSC component and the species with significant effects

are displayed in parentheses.

Chapter 3

72

Fig. 3.4 Mean concentrations of non-structural carbohydrates (NSC) as a sum of soluble carbohydrates (black bars) and starch (white

bars) in three plant parts (leaves, stems and roots) of seedlings of five treeline tree species. Seedlings were grown in a field experiment

with manipulations of vegetation cover and microclimate in the French Alps at 2100 m a.s.l. and samples were taken in two

subsequent seasons: at the end of the first growing season (September 2013) and at the end of the first winter (May 2014). BG = Bare

ground (no vegetation cover), IV = Intermediate vegetation cover, FV = Full vegetation cover, Sh = Shading (no vegetation cover).

Shown are means of three seedlings ± SD for total NSC only. For L. decidua in May 2014, FV, only one seedling could be sampled.

Leave

sSte

mR

oo

ts

Treatments

Larix decidua Picea abies Pinus cembra Pinus uncinata Sorbus aucuparia

Sep May Sep Sep Sep SepMay May May May

NSC

co

nce

ntr

atio

ns

(mg

g-1d

w)

BG BG BG BG BG BG BG BG BG BG

H2


73

Carbohydrate concentrations among the conifers were significantly lower in

full vegetation (L. decidua and P. uncinata: all three NSC components, P. cembra:

starch; Fig. 3.4, Appendix Table S3.2). This effect was strongly linked with an

interaction of treatment and season, showing reduced carbohydrate concentrations in

full vegetation only at the end of the summer but not anymore after the winter half-

year (P. uncinata: soluble sugars, P. abies: starch, by trend also in L. decidua: soluble

sugars). In P. cembra the seasonality in treatment effects was even stronger, with

increased NSC concentrations after the winter in full vegetation, while in the other

treatments concentrations decreased during winter as in the other species, ultimately

yielding significantly higher NSC concentrations in the full vegetation treatment than

in bare ground in spring (Fig. 3.4, Appendix Table S3.2). An additional treatment

effect was observed in P. uncinata, in which NSC concentrations were significantly

higher in intermediate vegetation (and by trend in the shading treatment) than on

bare ground (Fig. 3.4, Appendix Table S3.2).

The interactions between seedling part and season reveal that carbohydrate

reserves in the conifers decreased over the winter half-year in leaves (L. decidua:

soluble sugars) or both leaves and stem (P. cembra and P. uncinata: starch), while root

reserves remained relatively stable. In P. abies carbohydrate concentrations of

seedlings parts did not differ with the season (Fig.3.4, Appendix Table S3.2).

Variation in carbohydrate concentrations in seedlings of the broadleaved S.

aucuparia generally differed from that in the conifers, with one exception: Reserve

accumulation (especially starch) was highest in stem and roots, in contrast to the

evergreen species but similar to the deciduous L. decidua (Fig. 3.4, Appendix Table

Chapter 3

74

S3.2). In S. aucuparia, the only treatment effect showed an interaction with seedling

parts, with a significantly lower soluble sugar concentration in leaves from the

intermediate vegetation treatment compared to leaves from the other three

treatments (Fig. 3.4, Appendix Table S3.2). While starch and NSC concentrations

decreased over the winter as in the conifers, soluble sugars were significantly higher

in spring (Fig. 3.4, Appendix Table S3.2).

DISCUSSION

The influence of neighbouring herbaceous vegetation on early establishment of

treeline trees revolves around the importance of two opposed interactions –

competition and facilitation. While evidence of both interaction types has been

previously documented at different treeline sites (Moir et al. 1999; Germino et al.

2002; Dullinger et al. 2003; Maher et al. 2005), our study suggests that negative

impacts dominate. We found general competition effects of herbaceous vegetation on

tree seedlings as well as species-specific susceptibilities to combinations of

competition and indirect vegetation effects via microclimate or pathogens.

Competition is not evident across all seasons, however, and seasonally the outcome

may even switch to facilitation: in autumn / winter, evergreen species may benefit

from the protection by the remains of high vegetation to allow increased carbon

gains than in sites with low or no vegetation cover. The net effect in terms of growth

and survival is still generally negative for the first year, but this autumn window to

improve the carbon balance may be of utmost importance for the interaction between

trees and the alpine vegetation in determining the long-term tree establishment

success.


75

Competition dominates the interaction between tree seedlings and neighbouring

vegetation

Seedlings of the five focal tree species generally produced less biomass in plots with

neighbouring herbaceous vegetation (FV and IV) compared to plots without it (BG

and Sh; Fig. 3.3). Compared with the initial biomass, this means no biomass gain

with vegetation versus an approximate doubling without it. The only common

characteristic within these two pairs of treatments being the presence or absence of

neighbouring vegetation, competition rather than microclimate arguably causes this

result. This pattern further points towards belowground rather than aboveground

competition for light in determining growth, because in most species biomass in the

shading treatment – reducing light availability without neighbouring vegetation –

was similar to the bare ground treatment.

The observed dominance of negative vegetation effects is in line with

numerous studies in forestry on the detrimental effect of grass cover on tree

seedlings (e.g. Sims and Mueller-Dombois 1968; Nambiar 1990; Ellis and Pennington

1992), with the general observation that belowground competition for water and

nutrients prevails over competition for light in nutrient-poor environments (Coomes

and Grubb 2000; Strand et al. 2006; Bloor et al. 2008; Axelsson et al. 2014). Given i)

functional differences such as dense root mats of grassy vegetation versus shallow,

simple root system of young seedlings (Casper and Jackson 1997) and ii) nitrogen

being particularly limiting in cold ecosystems such as treelines (Thébault et al. 2014),

competition for soil resources may outweigh the potentially facilitative effects of

neighbouring herbaceous vegetation as documented at different treeline sites (Moir

et al. 1999; Dullinger et al. 2003). The contrast of our results with reports of

Chapter 3

76

exclusively positive interactions between treeline tree seedlings and grassland

vegetation (Germino et al. 2002; Maher et al. 2005) are probably related to differences

in the structure of the alpine plant communities in the respective study areas and /

or to different susceptibilities of the studied tree species. This is supported by the

findings of Bansal et al. (2011) using a gradient of herb cover, along which they

detected evidence for above- and belowground competition in treeline tree seedlings,

in line with our results, primarily under full herb cover, while survival of their two

study species was affected contrastingly by intermediate herb cover. Furthermore,

the type of response parameter measured might lead to contrasting results, since a

focus on stress effects (e.g. photoinhibition, mortality) may over-emphasize the

positive aspects of the interaction without revealing resource competition. Finally,

locally differing resource-related abiotic conditions (soil types, precipitation) as well

as alpine plant species composition can add further variation. For example, the

presence of legumes is known to enhance soil nitrogen (Li et al. 2015; Bowman et al.

1996), which could result in reduced competition for soil resources in favour of

facilitative effects.

In addition to these general patterns, there were several responses reflecting

species-specific negative impacts of neighbouring vegetation and / or microclimate.

First, P. cembra was the only species to show a lower biomass production in all

vegetation- or roof-covered plots compared to bare ground, which may be due to the

moister conditions in these treatments. A negative impact of soil moisture on early

establishment of this species has been shown previously (Loranger et al. 2016,

submitted), which agrees with its distribution at rather dry sites (e.g. steep slopes;

convex, exposed topographies; Brändli 1998; Didier 2001). Nevertheless, survival of


77

this species was high in all treatments and NSC concentrations were reduced only in

full vegetation cover at the end of the summer, indicating a low sensitivity to the

presence or absence of vegetation cover. This response might be enabled by the large

and rich seeds of this species, generally characterized as competition-intolerant,

providing the seedlings with long lasting reserves at least during the first growing

season (Ulber et al. 2004). Also P. abies appeared relatively insensitive to varying

vegetation cover, since we found no significant treatment effect for survival or NSC

concentrations. This may explain this species’ capacity to invade abandoned

subalpine mountain pastures (Bolli et al. 2007), where herb cover tends to be well

developed.

The most sensitive species were L. decidua and P. uncinata, in which full

vegetation cover resulted in strong reductions in survival, growth and total NSC

concentrations (Fig. 3.3). These two species are characterized as heliophiles,

regenerating on bare or sparsely vegetated ground (Rameau et al. 1993; Didier 2001;

Batllori and Camarero 2009), which matches their generally negative responses to full

vegetation. Interestingly, however, while growth was also reduced in intermediate

vegetation, there was no apparent limitation for NSC accumulation in this treatment.

This indicates that growth was limited through belowground competition, but

photosynthesis was not, due to sufficient light availability. Similarly interesting, the

shading treatment affected neither survival, growth nor NSC concentrations of both

these species (L. decidua and P. uncinata; Fig. 3.3). A possible explanation is that 30 %

light availability in the shading treatment (compared to 20 % under full vegetation)

may be above the necessary threshold to obtain a positive carbon balance, and thus

to support growth and survival. However, this remains difficult to confirm without

Chapter 3

78

an independent light availability gradient. Doubtlessly, these two species can tolerate

relatively shady conditions as long as belowground competition is absent. Since

seedling survival was negatively affected only by the full vegetation treatment, it can

be concluded that both below- and aboveground competition are needed to decrease

the seedling survival of these species. Notably, seedlings of these species (L. decidua

and P. uncinata) dying in full vegetation were always at least partly covered by

mould fungi. These seedlings probably experienced the highest intensities of

competition (i.e. above- and belowground). This can result in morphologically and

physiologically weakened individuals, resulting in a high susceptibility to pathogen

infections (Seiwa 1998), which is a primary cause of tree seedling mortality

(Yamazaki et al. 2009). Furthermore, pathogens might have been promoted in full

vegetation due to the humid and cool microclimate and close contact to neighbouring

plants.

In S. aucuparia, growth was also strongly reduced under both vegetation

treatments, but NSC concentrations were not. Seedlings of S. aucuparia are known to

be shade-tolerant (Raspé et al. 2000). Accordingly, carbon storage appeared to be

actively maintained at the expense of growth, a strategy associated with shade-

tolerance (Kobe 1997). Similarly to the previous two species, survival of S. aucuparia

was negatively affected by the combination of above- and belowground competition.

While in this case there was no visible cause of seedling mortality, the results are also

in accordance with the species’ regeneration niche in forest understorey (Zywiec and

Ledwoń 2008). There, light availability is low, but dense herbaceous vegetation is

generally absent, preventing the limiting combination of intense belowground


79

competition and shaded conditions experienced by seedlings in our full vegetation

treatment.

The preceding discussion highlights how competition (here belowground) can

alter the range of abiotic conditions that a species can tolerate, i.e. we show that the

fundamental niche (versus realized niche; see Hutchinson 1957) of L. decidua, P.

uncinata and S. aucuparia spans over a larger range of light conditions than expected

from their observed distribution. Indeed, L. decidua and P. uncinata – considered

heliophile species – can establish well in shaded conditions without belowground

competition (shading treatment), while S. aucuparia – considered shade-tolerant and

regenerating mainly in the understorey – does well in strong light conditions but not

in shaded conditions with competition. This strongly suggests that competition

causes their restricted observed realized niche.

Seasonal shifts from competition to facilitation depend on the leaf functional type

The negative impact of full vegetation cover on carbohydrate accumulation in

evergreen conifers was highly seasonal. Carbohydrate concentrations in this

treatment were lowest at the end of the summer, but after the winter half-year this

negative effect had disappeared in evergreen seedlings (P. abies, P. uncinata), or even

became positive in P. cembra (Figs 3.3-3.4). Hence, while reserves generally decreased

during winter, they increased for evergreen seedlings in the full vegetation treatment,

possibly due to a release from competition in this period. During summer, there was

intense competition by high and dense vegetation as indicated by the low biomass

increment of seedlings. At the end of the summer, at the time of the first sampling,

the aboveground parts of grasses and herbs started to die back. Since a permanent

Chapter 3

80

snow cover rarely develops at this site before December, seedlings with evergreen

leaves had a considerable period to be photosynthetically active. However,

transitional seasons such as autumn are particularly stressful, as frequent

combinations of cold temperatures and high irradiation can cause photoinhibition

(Germino and Smith 1999). Seedlings in the full vegetation treatment could now

benefit from partial shade in the remaining dead vegetation matrix and buffered

temperature conditions (Germino et al. 2002) and maintain or even increase their

carbohydrate reserves, while photosynthetic activity in the other treatments was

limited. To our knowledge, such a seasonal switch from competition to facilitation

has never been shown before for treeline tree species (but see Kikvidze et al. 2006 and

Venn et al. 2009 for examples in subalpine, herbaceous vegetation). These findings

demonstrate the dynamic potential of biotic interactions, even over short periods of

time, and their important influence on the early establishment of trees at their

elevational limit.

Unsurprisingly, the deciduous seedlings of S. aucuparia could not benefit from

the release of competition after the retreat of the herbaceous vegetation.

Characteristic for deciduous trees, they showed a large build-up of starch reserves in

stem and roots at the end of the summer, which then decreased in all treatments until

the end of the winter, probably most strongly after resuming activity in spring, i.e. at

bud break, shortly before our sampling (Kozlowski 1992). Seedlings of the deciduous

L. decidua are evergreen during their first years and show an intermediate response.

In this regard, the tendency of a seasonal vegetation effect indicates a certain benefit

from the extended productive season. This might contribute to advantages of the


81

ontogenetic switch of leaf types in this species as a strategy to alleviate competition

for small seedlings.

Our results show that herbaceous vegetation, which is dominant at and above

many treeline sites, exerts an important, mostly limiting impact on young tree

seedlings. This interaction can change depending on season and leaf type of the

seedlings, with a switch to facilitation leading to increased carbohydrate reserves in

the following spring. For most species this increase compensated for the lower NSC

concentrations due to competition in the previous summer, while for P. cembra it

even resulted in a net increase. It is, however, currently unclear whether these

increased reserves can sustain seedling growth and survival in dense herbaceous

vegetation over longer periods of time and whether the growth disadvantage from

the first year can be compensated.

In contrast to the clear responses in seedling performance, the germination

response was not affected by vegetation cover, showing that the abiotic conditions

remain within the range required by this earliest life-stage. Nevertheless, our seeds

were sown directly in the soil, whereas free-falling seeds will often be prevented

from reaching the soil by dense vegetation cover, undoubtedly reducing germination

success.

Although early seedling establishment is only the first step to successful

regeneration, our results suggest that tree establishment at and above alpine treelines

with dense herbaceous vegetation might be facilitated by disturbances creating sites

without vegetation cover. Indeed, such a relationship of seedling establishment with

exposed mineral soil caused by geomorphological processes or animal activities has

Chapter 3

82

been documented before (Malanson et al. 2009 and references therein) and might be

particularly important for species sensitive to dense vegetation cover, such as L.

decidua and P. uncinata. At the same time, without ground disturbances dense

vegetation cover may impede regeneration and explain treelines that are stable or

respond only slowly to warming temperatures. While these findings are an

important contribution towards an explanation of local treeline patterns and

dynamics, long-term studies including at least two growing seasons are urgently

needed to clarify if the observed seasonal facilitation results in a longer-term net

facilitation. Such studies should also include different life-stages and tree species to

take into account the dynamic potential of plant-plant interactions at a distribution

boundary.

ACKNOWLEDGEMENTS

We are very grateful to Serge Aubert, former director of the Lautaret Alpine

Botanical Garden and the affiliated research station Joseph Fourier, for enabling this

study by allowing us to use the experimental space and facilities of the station, and

thank the gardener team for support with any technical problem. We also thank

several field assistants, especially Jasmin Baruck, Gesa Pries, Verena Schenk and Eric

Thurm, for their help in installing, maintaining and monitoring the experiment. The

study was funded by the German Research Foundation (DFG, BA 3843/5-1&2).


83

APPENDIX

Table S3.1. Summary of paired t-tests comparing microclimatic conditions between

treatments. Compared were daily minimum, maximum and mean relative air

humidity (rh) in treatments with three different levels of vegetation cover measured

in a seedling transplant experiment in the French Alps at 2100 m.

Treatment

comparisons Block 1 Block 2 Block 3

FV vs. IV

rh min d = 12.1, p < 0.001 d = 12.5, p < 0.001 d = 4.1, p = 0.19

rh max d = 0.4, p = 0.06 d = 1.0, p = 0.06 d = - 0.5, p = 0.52

rh mean d = 6.1, p < 0.001 d = 9.6, p < 0.01 d = - 2.7, p = 0.09

FV vs. BG

rh min d = 20.1, p < 0.001 d = 23.0, p < 0.001 d = 7.0, p = 0.09

rh max d = 0.6, p = 0.08 d = -0.2, p = 0.36 d = 0.1, p = 0.51

rh mean d = 9.1, p < 0.001 d = 9.2, p < 0.001 d = 4.1, p < 0.001

Rows show results of treatment pairwise comparisons of minimum, maximum and mean values

giving the mean of the differences (d, %) and the p-value with significant effects given in bold. BG =

Bare ground (no vegetation cover), IV = Intermediate vegetation cover, FV = Full vegetation cover.

Chapter 3

84

Table S3.2. Summary of ANCOVA models testing the differences in non-structural carbohydrates (NSC) and of its two components

(soluble carbohydrates and starch) between treatments, plant parts and seasons. Seedlings of five treeline tree species were sampled at

the end of the first growing season and the end of the first winter 2013/2014 in a field experiment in the French Alps at 2100 m.

Soluble Carbohydrates Starch NSC

F p Effect F p Effect F p Effect

Larix decidua

Treatment 6.02 0.001 BGa, IVa, FVb, Sha 6.64 <0.001 BGab, IVa, FVb, Sha 15.37 <0.001 BGa, IVa, FVb, Sha

Part 2.13 0.13 … 13.14 <0.001 Lb, Sa, Ra 4.1 0.02 Lb, Sab, Ra

Season 1.18 0.28 … 9.23 <0.01 13a, 14b 7.33 <0.001 13a, 14b

Treat:Part 0.83 0.55 … 1.98 0.09 … 1.05 0.41 …

Treat:Season 1.63 0.19 … 1.88 0.15 … 3.28 0.03 …

Part:Season 11.35 <0.001 L13>{R13, S13,

L14}

R14>L14

0.17 0.85 … 8.38 <0.001 L14<{R14, S14, L13}

Picea abies

Treatment 1.68 0.18 … 1.48 0.23 … 1.67 0.18 …

Part 2.36 0.10 … 10.69 <0.001 Lb, Sb, Ra 1.99 0.15 …

Season 0.05 0.83 … 0.05 0.82 … <0.01 0.95 …

Treat:Part 0.73 0.63 … 0.41 0.87 … 0.49 0.81 …

Treat:Season 0.43 0.74 … 6.33 <0.001 Sh13>{FV13, Sh14} 2.34 0.08 …

Part:Season 0.09 0.91 … 2.25 0.11 … 0.72 0.49 …

Pinus cembra

Treatment 1.96 0.13 … 10.17 <0.001 BGa, IVa, FVb, Sha 2.03 0.12 …

Part 0.67 0.52 … 43.13 <0.001 Lb, Sb, Ra 10.70 <0.001 Lb, Sb, Ra

Season 0.57 0.45 … 10.31 <0.01 13a, 14b 0.31 0.58 …

Treat:Part 0.89 0.51 … 2.01 0.08 … 1.16 0.34 …

Treat:Season 5.59 <0.01 FV14>{BG14,

FV13}

6.26 <0.001 BG14< BG13 10.00 <0.001 BG14<FV14

FV13<{ BG13, IV13,

FV14}

Part:Season 1.74 0.19 … 10.98 <0.001 [L, S]13>[L, S]14 0.08 0.93 …


85

Pinus uncinata

Treatment 8.26 <0.001 BGab, IVa, FVb, Sha 97.28 <0.001 BGa, IVa, FVb, Sha 18.75 <0.001 BGb, IVa, FVc, Shab

Part 11.03 <0.001 La, Sb, Ra 28.57 <0.001 Lb, Sb, Ra 15.08 <0.001 La, Sb, Ra

Season 0.75 0.39 … 27.05 <0.001 13a, 14b 2.71 0.11 …

Treat:Part 1.36 0.25 … 1.68 0.15 … 1.00 0.43 …

Treat:Season 5.28 <0.01 FV13<(BG, Sh,

IV)13

5.92 <0.01 BG13>BG14;

Sh13>Sh14

6.94 <0.001 BG14<{ IV14, BG13}

Part:Season 1.60 0.21 … 33.19 <0.001 L13>L14; S13>S14

R14>(L14, S14)

7.75 0.001 R14>{L14, S14}

Sorbus aucuparia

Treatment 2.71 0.05 … 2.64 0.06 … 0.53 0.66 …

Part 0.77 0.47 … 16.68 <0.001 Lb, Sa, Ra 6.96 <0.01 Lb, Sa, Ra

Season 7.14 0.01 13b, 14a 102.4 <0.001 13a, 14b 31.10 <0.001 13a, 14b

Treat:Part 2.88 0.02 IV-L<{FV-L, Sh-L} 0.45 0.84 … 1.25 0.30 …

Treat:Season 1.25 0.30 … 0.69 0.56 … 0.80 0.50 …

Part:Season 36.58 <0.001 L14<{ L13, S14,

R14}

R13<{R14, L13}

S13<{S14, L13}

5.42 <0.01 L13>L14

S13>{L13, S14}

R13>{L13, R14}

2.17 0.12 …

Shown are the main effects treatment, seedling part and season as well as their two- and three-way-interactions for the three response variables soluble

carbohydrates, starch and total NSC concentrations of each species reporting F- and p-values. Significant effects (p<0.05) are given in bold and italics, marginally

significant effects (p<0.1) only in italics. Notes - Treatment: BG = Bare ground (no vegetation cover), IV = intermediate vegetation, FV = full vegetation, Sh =

Shading (no vegetation cover). Part: L = leaves, S = stem, R = roots. Season: 13 = September 2013, 14 = May 2014. Significant differences between treatments,

parts, or seasons are represented with different letters in superscript, with the first letter in alphabetical order representing higher carbohydrate concentrations (a

> b).

87

CHAPTER 4

A COOL EXPERIMENTAL APPROACH TO EXPLAIN ELEVATIONAL

TREELINES, BUT CAN IT EXPLAIN THEM?

Maaike Y. Bader, Hannah Loranger, Gerhard Zotz

Published in American Journal of Botany

ABSTRACT

At alpine treeline, trees give way to low-stature alpine vegetation. The main reason

may be that tree canopies warm up less in the sun and experience lower average

temperatures than alpine vegetation. Low growth temperatures limit tissue

formation more than carbon gain, but whether this mechanism universally

determines potential treeline elevations is the subject of debate. To study low-

temperature limitation in two contrasting treeline tree species, Fajardo and Piper (

Chapter 4

88

American Journal of Botany 101: 788–795) grew potted seedlings at ground level or

suspended at tree-canopy height (2 m), introducing a promising experimental

method for studying the effects of alpine-vegetation and tree-canopy microclimates

on tree growth. On the basis of this experiment, the authors concluded that lower

temperatures at 2 m caused carbon limitation in one of the species and that treeline-

forming mechanisms may thus be taxon-dependent. Here we contest that this

important conclusion can be drawn based on the presented experiment, because of

confounding effects of extreme root-zone temperature fluctuations and potential

drought conditions. To interpret the results of this elegant experiment without

logistically challenging technical modifications and to better understand how low

temperature leads to treeline formation, studies on effects of fluctuating vs. stable

temperatures are badly needed. Other treeline research priorities are interactions

between temperature and other climatic factors and differences in microclimate

between tree canopies with contrasting morphology and physiology. In spite of our

criticism of this particular study, we agree that the development of a universal

treeline theory should include continuing explorations of taxon-specific treeline-

forming mechanisms.

Key words: alpine treeline; carbon balance; ecophysiology; Fuscospora; growth

limitation; methodology; micrometeorology; Nothofagus; Pinus; timberline

Methodological Approaches

89

INTRODUCTION

In a recent paper in the American Journal of Botany , Fajardo and Piper (2014)

presented an experimental approach that compares tree seedling performance in

conditions typical for low-stature alpine vegetation and conditions representing tree

canopy conditions at treeline. The rationale of this approach was that treelines are

assumed to form because tree canopies cannot warm up as much as low-stature

alpine vegetation can, due to a stronger coupling to atmospheric conditions (Körner

1998). The lower temperatures experienced by a tree canopy do not allow growth,

either because carbon assimilation is insufficient (the source-limitation or carbon-

limitation hypothesis), or because growth processes (e.g., cell division or

lignification) are directly impaired (the sink-limitation or growth-limitation

hypothesis). Although most results of recent research, in particular on elevational

patterns of non-structural carbohydrate contents in trees (summarized by Hoch and

Körner 2012) and on low-temperature limits to wood formation (e.g. Rossi et al.

2008), support the growth limitation hypothesis, some findings seem to point at

carbon limitation (Wiley and Helliker 2012; Dawes et al. 2013; though see Palacio et

al. 2014). This issue is thus not generally resolved.

The new approach used by Fajardo and Piper (2014) aimed to study low-

temperature limitation in trees experimentally by placing individuals of similar size

and developmental stage in temperature regimes usually experienced by individuals

of very different sizes, thereby avoiding any confusion of microclimatic effects with

ontogenetic and size effects. Even though tree seedlings were used, this approach

was not aimed at studying seedling performance as such but used seedlings as a

Chapter 4

90

small-version model of a taller tree to study responses of tree growth to

microclimatic conditions.

In the experiment, one of the two species tested, Nothofagus pumilio , showed a

decrease in growth and non-structural carbohydrate (NSC) reserves when suspended

at 2 m above the ground, where mean temperatures were lower than at ground level.

The other species, Pinus contorta, showed no response. The authors concluded that

their study is the first unambiguous test of the mechanism behind growth limitations

at treeline elevation, with evidence for differences between treeline-forming taxa.

Such an important finding deserves thorough scrutiny. Here, we argue that due to

their particular experimental method, the observed carbon limitation may be due to

other factors than low mean temperature and does not represent strong and

unambiguous evidence for a treeline-specific phenomenon.

However, we do agree with the authors that potential taxon-specific, growth-

limiting mechanisms at treeline should be seriously considered when further

developing alpine-treeline theory. First, because we should continue to question

whether there is really one mechanism explaining this low-temperature life-form

boundary or whether equally valid representatives of this life-form are constrained

differently. And second, because only by systematically exploring the variation

beyond a universal treeline-forming mechanism can we hope to understand and

predict treeline elevations in real landscapes.


91

CONFOUNDING TEMPERATURE CONDITIONS

It is an excellent idea to test differences in plant performance under low-vegetation-

and tree-canopy-temperature regimes independent of ontogeny and size. Growing

seedlings at ground level and at canopy level (but outside an actual canopy) to mimic

the different levels of atmospheric coupling is a highly promising approach.

However, there is one fundamental problem: soil temperatures in suspended pots

fluctuate very strongly. They fluctuate much more strongly than soil temperatures at

ground level and also more strongly than air temperatures around the suspended

pots ( but less than air temperatures at ground level, see Fig. 2 of Fajardo and Piper,

2014). In the experiment of Fajardo and Piper (2014), suspended seedlings thus

experienced similar temperature fluctuations in roots (3–10 ° C) and shoots (3–8 ° C),

whereas ground level seedlings experienced low root-zone fluctuations (6–8 ° C) and

high shoot-zone fluctuations (2–14 ° C).

Such strong temperature fluctuations in suspended pots clearly do not mimic

root-zone conditions below a tree canopy, which are very stable because of the large

soil volume and the shade provided by the canopy (Körner and Paulsen 2004). In the

experiment of Fajardo and Piper (2014), the average temperature in the suspended

pots was about 1 K lower than in ground-level pots. In that sense, the suspended

pots did mimic one aspect of soil conditions under a tree canopy, which are nearly

always cooler during the growing season than under nearby alpine vegetation

(Bendix and Rafiqpoor 2001; Bader et al. 2007; Körner 2012). The crucial question

here is, however, what do these mean temperatures mean physiologically? Biological

rates respond nonlinearly to temperature, so that at a given mean temperature,

Chapter 4

92

fluctuating temperatures should lead to different mean biological rates than constant

temperatures. On the other hand, mean growing season temperatures are the most

consistent thermal parameter at treelines worldwide (Körner and Paulsen 2004;

Cieraad et al. 2014; Paulsen and Körner 2014) and experimental results from Hoch

and Körner (2009) showed similar growth rates in two conifer species at constant and

variable temperatures. Although these observations still await a physiological

explanation, they suggest that mean temperature really has a biological meaning. The

lower mean root-zone temperature in the suspended pots could thus rightfully be

expected to slow seedling growth. However, as long as it cannot be excluded that the

strong temperature fluctuations contributed to the observed differences in seedling

performance, the observed effect cannot be unequivocally attributed to the different

mean temperatures.

How likely is it that these fluctuations were really a problem? The answer

depends on how temperature differences between root and shoot and on how

temperature fluctuations in general affect seedling physiology. Although these seem

two very basic biological questions, there is surprisingly little information to guide

an answer (Pregitzer et al. 2000). Opposite effects of root and shoot temperature on

plant nutrient status (Weih and Karlsson 2001) and biomass allocation patterns

(Larigauderie et al. 1991) suggest that temperature differences between roots and

shoots, aside from their absolute temperatures, can affect whole-plant performance.

However, we did not find experiments focusing explicitly on this question.

Temperature-fluctuation-effects on plant growth are hardly studied either. In an

experiment addressing this for trees at low mean temperatures, conifer seedlings

grew similarly well under constant or variable (ca. 6 K amplitude around 6 ° C or 12 °


93

C) daily and seasonal temperature regimes, with only slight positive effects of

fluctuations for Larix decidua but not for Pinus mugo (Hoch and Körner 2009). In

contrast, in the only other experiment with trees that we could find, temperature

variability (5 or 10 K amplitude around 23 ° C, with a 5-d fluctuation) clearly affected

growth and root to shoot ratios in poplar ( Populus deltoids × nigra ) cuttings, with

positive effects of fluctuations at the intermediate but not at the greater amplitude

(Cerasoli et al. 2014). Obviously, this limited and ambiguous evidence does not allow

generalizations. Additional experiments addressing temperature-fluctuation effects

on tree growth, addressing differences between species or functional types, the mean

temperature and the amplitude of the fluctuations, are clearly desirable. Apart from

identifying potential artifacts in experiments like that of Fajardo and Piper (2014),

such studies could greatly contribute to understanding the physiological meaning of

different temperature parameters for treeline formation.

CONFOUNDING MOISTURE CONDITIONS

There is a second concern: how severe was drought stress in these suspended plots,

and did this stress affect the results? Seedlings were only watered during the first

month of the growing season. Even though precipitation during the growing season

was ca. 500 mm at this Patagonian site (Fajardo and Piper 2014), this statistic does not

preclude that rainless periods were frequent and that the soil in the pots dried out

repeatedly during these periods, especially in the suspended pots. If P. contorta is less

sensitive to drought than N. pumilio , this sensitivity could explain the species-

specific responses. The documented carbon limitation of N. pumilio (reduced growth

Chapter 4

94

associated with low NSC contents) could thus be caused by water stress and not by

low temperature.

The authors argued that water shortage should not have affected N. pumilio

seedlings, based on results of earlier experiments where seedlings did not respond to

watering in ambient and warmed conditions (Piper et al. 2013). However, not

requiring extra watering in full-soil conditions does not imply drought tolerance in

suspended pots. Playing devil’s advocate, one could even argue that this earlier

experiment did show a trend, though not significant, to higher growth rates under

watering (Piper et al. 2013). Similar negative effects of drought (as a result of

experimental warming) have been observed at treeline in New Zealand for seedlings

of Fuscospora cliffortioides ( previously Nothofagus solandri var. cliffortioides, Heenan

and Smissen 2013) (Melanie Harsch, University of Washington, Seattle, personal

communication) and in North America for recently germinated Pinus flexilis

seedlings (Moyes et al. 2013). In the contested experiment by Fajardo and Piper

(2014), the higher mortality (significant when combining both species) in the

suspended pots also suggests a stress factor confounding or aggravating a potential

temperature effect.

To conclude, the study by Fajardo and Piper (2014) does not, because it could

not, provide unambiguous support for low temperature- induced carbon limitation

in N. pumilio. Because adult trees of the same species show no decrease in NSC with

elevation and thus do not appear carbon limited (Fajardo et al. 2011), the

fundamentally different growth-limitation for Nothofagus compared with other

treeline trees remains to be shown.


95

ALTERNATIVE EXPERIMENTAL SETUPS

Finding problems in an experiment is easy. More difficult and more useful is

proposing solutions. Understanding how low temperatures limit tree growth is not

only of fundamental biological interest but also affects model predictions of tree

growth in a warmer and CO 2 - richer future. Temperature effects can be studied in

isolation in fully controlled conditions in growing chambers (e.g. Hoch and Körner

2009). However, it is very difficult to mimic the typical treeline combination of high

radiation and low air temperature. For the question of carbon vs. growth limitation,

in particular, results from such experiments will be hard to translate to the real

world.

In the field, one option for studying temperature effects on treeline tree

growth is to warm existing tree canopies. However, this technically and logistically

very challenging manipulation is practicable only for small sections such as branches

(Lenz et al. 2013) or buds (Petit et al. 2011). Such sectional studies can yield important

information about local growth processes but cannot contrast carbon vs. growth

limitation directly, because these involve microclimatic effects on whole-tree carbon

gain and use.

Using seedlings as model tissue for tree performance is a sensible alternative.

Compared with adult trees, they allow for better replication and faster responses and

a distinction of microclimatic from ontogenetic and size effects. Suspended seedlings

experience tree canopy conditions, but the results may be confounded by soil

temperature fluctuations not seen in natural soil and by uncontrolled soil moisture.

The moisture issue is relatively easy to solve (disregarding logistic difficulties for

Chapter 4

96

maintaining an irrigation system at treeline). Whether temperature fluctuations are a

problem remains to be investigated, as discussed above. As long as this is unclear,

temperature needs to be controlled. Temperature fluctuations in suspended pots can

be reduced by using reflective material, insulation, and ventilation layers around the

pots. However, in trials related to a somewhat similar experiment, we were unable to

design pots that can be suspended in air while maintaining the stable temperature

conditions typical for full soil, let alone for full soil under a tree canopy (see

Appendix S4.1). We tried this in northern Germany (ca. 53 ° N) in early spring, with

relatively mild sunshine loads. At the alpine treeline in the middle of summer and at

lower latitudes, we would expect the problem to be worse. Active temperature

control using flowing water around the pots could work well from the physical point

of view, though from the practical point of view such a system would be very

challenging to install at most treeline sites.

Assuming that root-zone temperatures can be controlled, the question remains

which temperature regimes would allow an informative comparison between

canopy-level seedlings and ground-level seedlings. Seedlings can be used as models

to study tree growth in two distinct ways: as a model for tree tissue in general or as a

small-version model of a tall tree. These approaches require explicit assumptions

about temperature effects and require different temperature regimes. Using seedlings

as a model for tree tissue in general seems reasonable for some questions because

roots and shoots have similar temperature thresholds for growth (references in Rossi

et al. 2008) and both need to be warm enough for a plant to grow (Körner and Hoch

2006). For such a model, roots and shoots should ideally experience the same

temperature regimes. Therefore, root-zone temperatures in suspended and ground-


97

level pots should be regulated to follow the respective air temperatures (and

irrigation adjusted accordingly).

The alternative, using seedlings as a small-version model for a tree, assumes

that whole-plant physiology is affected differently by root than by shoot temperature

(Larigauderie et al. 1991; Weih and Karlsson 2001) and that this is similar for

seedlings and adult trees. Such a small-version tree model is what Fajardo and Piper

(2014) had in mind for their experiment, which “aimed to mimic the low temperature

effects on meristematic shoot and root tissues of a taller tree and, eventually, the

temperature effects on the tree’s C balance as a whole”. For such a model,

temperatures in the pots should follow the respective root-zone temperatures in

alpine vegetation (ground-level pots) and below trees (suspended pots). Thus, only

suspended pots would need regulation (assuming the ground-level pots are buried

in the ground and temperatures there are naturally representative), and this

regulation could be based on measured soil temperatures under a nearby tree

canopy. Alternatively, if the question is focused on aboveground temperatures only,

root-zone temperatures could be similar in both treatments, i.e., could be regulated in

suspended pots to follow the temperature in the ground-level pots.

A common approach to manipulate temperature for seedlings at treeline is

warming, either passively by using transparent roofs (Germino and Smith 1999) or

open-top chambers (e.g. Danby and Hik 2007; Xu et al. 2012), or actively by using

infrared lamps (Moyes et al. 2013). All these methods are useful for studying thermal

constraints for seedlings and for mimicking future climate warming, keeping in mind

restrictions inherent to the different types of warming. For mimicking tree canopy

Chapter 4

98

conditions, however, seedlings need to be cooled, which is more challenging than

warming. Other than suspending seedlings in the air, options would be planting

under the tree canopy or with artificial cover (but there will be confounding effects of

shade), planting at higher elevations (but this option implies cooling both day and

night, in contrast to tree canopies, which are warmer at night than low vegetation), or

active cooling. Active cooling could be achieved via radiative cooling (e.g., using

peltier or other electric cooling elements near the plants) or via convective cooling

(using ventilators, supplying air from outside the soil–alpine vegetation boundary

layer). As an imitation of convective cooling (and at night: warming) of the tree

canopy, the ventilator option seems by far the better choice. This approach, on a

larger scale using wind machines, is commonly used in horticulture to prevent

radiation frosts (Perry 1998). As in the suspended-pot experiment presented by

Fajardo and Piper (2014) , such a setup would have to control for soil moisture

differences. Belowground temperatures could either be left to equilibrate with the air

(the easier option), or aboveground ventilation could be accompanied by

temperature regulation of the root zone to mimic soil conditions below a tree canopy,

e.g., using electric temperature elements as is sometimes used, though again usually

for heating only, in climate-change experiments (Melillo et al. 2002).

In all of these setups, apart from accounting for potential microclimatic

artifacts, the assumed role of the seedlings as models for tree growth should be made

explicit and validated before translating the results obtained with the seedling

models to adult trees.


99

Comparing taxa regarding canopy–microclimate effects on tree growth is only

one side of the story, however. To understand fully the role of microclimate in

limiting tree growth, differences in canopy microclimates and tissue temperatures

due to taxon-specific leaf and crown morphologies, transpiration rates and albedo

also need to be considered (Leuzinger and Körner 2007). Data addressing such

differences for treeline trees are rare (Körner 2012), though they would be relatively

easy to obtain using data loggers. Such data would be another valuable contribution

toward understanding the mechanisms of treeline formation for different taxa and

functional tree types.

ALTERNATIVE TREELINE-FORMING MECHANISMS IN NOTHOFAGUS AND OTHER

GENERA

Nothofagus treelines have long been regarded as unusual, occurring at higher

temperatures than most northern-hemisphere treelines, which was explained by

genus-specific limitations (Körner and Paulsen 2004; Wardle 2008). However, several

recent studies suggest that mean temperatures in the growing season at these

treelines are actually quite comparable to those at other treelines (Mark et al. 2008;

Cieraad et al. 2014; Fajardo and Piper 2014). Another argument against genus-specific

limitations is presented in the study discussed here (Fajardo and Piper 2014), where

Pinus contorta seedlings did not outperform Nothofagus pumilio at 50 m above the

treeline. The authors argue that because of the lower mean temperatures in the

suspended pots, P. contorta should not outperform N. pumilio up to 330 m above the

treeline, although this argument disregards temperature extremes. However, in an

experiment in New Zealand started in the 1960s, Pinus contorta and other exotic

Chapter 4

100

conifers developed large stems up to 300 m above the local treeline, while saplings of

the native treeline species Fuscospora cliffortioides (Nothofagus solandri var.

cliffortioides), though surviving ca. 150 m above the treeline, have still not emerged

from the boxes that protected them as seedlings (Fig. 4.1). Clearly, Pinus is

outperforming this close relative of Nothofagus in this case (Wardle 1985). So the

question remains: do the Nothofagaceae form treelines for different reasons than

most other treeline tree families?

Fig. 4.1 A. Overview of the treeline tree-establishment experiment installed by Peter

Wardle in the 1960s in the Craigieburn Range, New Zealand (photos taken in March

2009 by M. Y. Bader). View down toward the treeline from the experimental garden

at 1450 m, ca. 150 m above the current treeline. Exotic tree species, many of which

have grown well at this elevation, have been cut down to prevent them from

spreading into the native ecosystem. Note the shrubby Fuscospora cliffortioides (≡

Nothofagus solandri var. cliffortioides), the local treeline tree species that appears to

have survived or established in the shelter of the now cut-down exotic trees. B.

Stump of Pinus contorta at 1450 m. The box is the remnant of an experimental 73%

shade treatment. The rosette plant in the box is the alpine species Aciphylla cf. aurea ,


101

unrelated to the experiment. C. Shade box with barely grown F. cliffortioides

apparently unable to escape these sheltered conditions (at 1450 m) (experiment and

results described in Wardle, 1985 , 2008).

Most treelines composed of Nothofagaceae are very abrupt boundaries from

tall, closed forest to low alpine vegetation, suggesting that limitations to

establishment outside the forest rather than limited growth determine the position of

these treelines (Harsch and Bader 2011). Above the natural treeline, Fuscospora

cliffortioides seedlings depend on shade and/or frost protection, which also suggests

that tree establishment and treeline advance are not limited by low growing

temperatures alone but by the interaction with stressors like frost, excess radiation,

and wind (Wardle 1993, 2008). Such stressors cannot offer a universal explanation for

treeline formation, as they vary strongly among treelines and can occur at any

elevation (Körner 2012). In the absence of such stressors and in the case of resistant

species, treeline elevations can be controlled by low temperature limitations to

growth, either via source or sink limitation. In all other cases, regional peculiarities

and “taxon-specific” treelines emerge (Harsch and Bader 2011). As these are arguably

the rule rather than the exception (e.g. Piper et al. 2006; Wardle 2008; Holtmeier 2009;

Malanson et al. 2011), we embrace the recommendation of Fajardo and Piper (2014)

to keep developing a universal theory for treeline formation including variation

between tree taxa and regional climates. Experiments to this end should address low-

temperature limitations to growth as well as interactions with other climatic factors

in all tree life stages. Preferably, such experiments should include several members of

different functional tree types to allow generalizations beyond taxon-specificity.

Chapter 4

102

CONCLUSION

The presented experimental setup in Fajardo and Piper (2014) is a low-technology

approach for studying a fundamental question in functional plant biology: what

causes treelines? Though elegant, the experiment suffers from a potentially

disqualifying practical problem: the strong fluctuations in soil temperature (and most

probably soil moisture) in the suspended pots. As long as the effect of such

fluctuations is unknown, they cannot be ignored and should be experimentally

controlled. A promising alternative is to keep seedlings at ground level and ventilate

them to mimic the atmospheric coupling found in tree canopies. These solutions

require an infrastructure that is not usually available at treeline, though solar panels

and local water sources could allow these setups even in remote sites. Apart from

technical solutions, we discussed conceptual solutions: how can we interpret the data

given the temperature and moisture fluctuations? To do this, soil moisture data

would be needed as well as a much better understanding of the effects of

temperature fluctuations on tree growth. At this stage, an unambiguous

interpretation of the results of Fajardo and Piper (2014) seems impossible. Nothofagus

pumilio seedlings are carbon-limited before they are growth-limited at the

temperature and moisture conditions in the suspended pots, but it is unresolved

whether the limitations are really due to the lower mean temperature or (1) to the

strongly fluctuating root-zone temperature regime, or (2) to moisture stress.

Although their experiment did not allow unambiguous conclusions about the causes

of alpine treeline formation, it provides excellent food for thought on further

experiments toward this goal.


103

APPENDIX

Fig. S4.1 A. Vertically adjustable planting-tube designed to test low-temperature

limitations of tree growth and vegetation microclimate on seedling performance by

growing same-aged tree seedlings at different heights within and above high alpine

grassland vegetation. An inner tube (3 cm planting core) is wrapped in 2 cm-thick

pipe insulation material and suspended in an outer reflective aluminium tube,

leaving a 1cm aeration layer through which air can circulate from the aeration holes

(top) via the open bottom. Note that the soil surface is covered with moss for

comparability of temperature measurements with the moss-covered ground. B.

Comparative temperature measurements between the soil in the planting tube at 100

cm height, the soil in the ground and the air at 100 cm height. Temperature sensors

were inserted 5 cm into the soil for soil temperature measurements and protected

from solar radiation by an aerated shield (not shown). Solid black line: soil

temperature in the ground, dotted black line: soil temperature in the planting tubes,

dotted grey line: air temperature.

Date

Tem

pe

ratu

re°C

A B

105

CHAPTER 5

SYNTHESIS

On the one hand, the studies presented in this thesis investigated experimentally

how abiotic and biotic environmental conditions affect the two earliest regeneration

stages of trees at the alpine treeline. Soil moisture was, in addition to temperature, a

critical factor for both germination and subsequent seedling establishment. The

interaction with neighbouring herbaceous vegetation affected only the seedling life-

stage and was principally negative, but could seasonally switch to facilitation for

evergreen tree species. While general patterns emerged for responses to both

microclimate and plant-plant-interactions, species-specific effects were more strongly

represented. On the other hand, we pointed out in a commentary the difficulty of

finding resourceful, novel approaches to test hypotheses and gain mechanistic

insights to explain a major vegetation boundary with an array of potentially limiting

and interacting factors, and explored the conceivable possibilities.

Chapter 5

106

LOCAL AND INTRINSIC FACTORS DRIVING THE REGENERATION RESPONSE

It is still under debate if an overarching explanation for treeline formation can be

found on a global scale, or if individual treelines, although ultimately limited by heat

deficiency, are each constrained by different sets and interactions of site- and taxon-

specific factors. The regeneration of treeline trees as a potential life-history bottleneck

plays an important role in this debate, and the presented studies show that early

establishment is importantly affected by local environmental factors such as

microclimate and biotic interactions as well as intrinsic factors such as life-stage

dependencies and tree species ecology.

In terms of environmental factors, we confirmed the well-established

importance of temperature for treelines in general (Rolland et al. 1998; Körner and

Paulsen 2004) and for early regeneration stages in particular (Germino and Smith

1999; Smith et al. 2003) by showing the expectedly negative impact of decreasing

temperatures on germination and early seedling establishment. We further found,

however, that temperature effects could not be decoupled from the effect of water

availability and that depending on the species, the importance of the latter

predominated (Chapter 2). While varying intensities of vegetation cover significantly

altered the microclimate regarding both temperature and water availability, the

results from this experiment suggest that direct negative and especially belowground

plant-plant interactions, i.e. competition, between tree seedlings and herbaceous

alpine vegetation prevail (Chapter 3). Both studies reveal a strong dependency of

early tree life-stages on soil resources such as water or nutrients, which can be linked

i) for germination to the requirement of high moisture conditions as external cue and

Synthesis

107

driver of physiological processes (Baskin and Baskin 2001) and ii) for seedlings to

their characteristically small size connected with only small amounts of productive

and storage tissue, simple rooting system and a consequently low resource-uptake or

storage capacity (Wang and Zwiazek 1999; Johnson et al. 2011).

The responses of the studied tree species to microclimatic manipulations, both

positive and negative, were generally consistent over the earliest life-stages of

germination and subsequent seedling establishment (Chapter 2). This concordant

pattern has important implications for the regeneration success: it reduces

restrictions caused by seed-seedling conflicts regarding the environmental factors

required for establishment, thereby mitigating the effect of the environmental filters

imposed on each life-stage and supporting that a safe site (sensu Harper 1977) for the

seed is also safe for the seedling (Schupp 1995). However, in a relatively stable

environment it can also be expected that limiting factors have accordingly a higher

impact on a regeneration process with concordant life-stage responses, whereas an

irregularly occurring stress factor may be temporarily decoupled from brief

susceptible stages such as germination (Shen et al. 2014). The fact that we did not

find concordant patterns over both early life-stages for interactions with the alpine

herbaceous vegetation, i.e. no effect on germination and important impacts on the

seedling stage (Chapter 3), can be related to two important conditions: i) due to the

reserves contained in the seed, the germination success is relatively independent

from external resources and thus much less susceptible to competition, which was

the dominant interaction affecting seedling performance, and ii) the abiotic

conditions, although modified by varying intensities of vegetation cover, were

apparently not altered beyond the range required by this earliest life-stage. However,

Chapter 5

108

since the presented studies were conducted over a relatively short period of time

with measurements taken primarily in the less stressful growing season and

germination was monitored from seeds sown directly in the ground, it remains

unclear how permanent versus variable environmental factors impact the overall

regeneration success and how vegetation cover affects the germination of free-falling

seeds.

The previous paragraphs show that general patterns emerged in both studies

investigating the impact of potential environmental limitations on early regeneration

responses, but nevertheless, species- or taxon-specific effects predominated. In terms

of abiotic environmental factors, the relative importance of the manipulated

microclimate variables temperature and moisture as well as the direction of effects

were highly idiosyncratic (Chapter 2). These patterns on the one hand fittingly

represent the characteristics of the species’ ecology: for example, the high importance

of moisture in comparison to temperature for both early establishment stages of L.

decidua, although as young seedling still evergreen, may reflect the low water use

efficiency of later deciduous life-stages (Matyssek 1986). On the other hand they may

relate specific limitations to the species’ geographical distribution, as germination in

P. uncinata, a species forming southern treelines such as in the Southern European

Alps and in the Pyrenees (Rameau et al. 1993; Batllori and Camarero 2009), was

particularly sensitive to low temperatures. In terms of interactions with the alpine

herbaceous vegetation, tree-species-specific negative responses were principally

caused by combinations of competition and indirect vegetation effects such as

microclimate and pathogens, which might decrease seedling performance through

physiological and morphological weakening followed by increased disease

Synthesis

109

susceptibility (Seiwa et al. 2008; Yamazaki et al. 2009). The conifer species however,

all evergreen in the early seedling stage, additionally showed the potential to

seasonally switch the interaction with high herbaceous vegetation from competition

to facilitation: between the dieback of the vegetation and the development of a

permanent snow cover they could benefit from an extended productive season in the

shelter of vegetation remains, thereby compensating for carbon losses due to

competition during the summer or even obtaining a positive carbon balance at the

end of the winter (Chapter 3). While previously shown for herbaceous alpine plant

communities (Kikvidze et al. 2006; Venn et al. 2009), such a seasonal switch in species

interactions has to our knowledge never been demonstrated for treeline trees before.

Thus, this taxon-specific response highlights the dynamic potential of biotic

interactions at a distribution boundary even over short periods of time. It remains

however still unclear if this seasonal benefit results in long-term net facilitation, or if

competitive effects of the herbaceous vegetation prevail.

The important and variable impact of the site-specific and intrinsic factors on

two critical stages of early establishment in our studies suggests a frequent restriction

of treeline tree regeneration before a temperature limit for growth is reached.

Germination and early seedling establishment are not sufficient, but indispensable

first steps towards the establishment of a mature tree. Although several subsequent

life-stages have to be completed before a tree contributes to the current tree line, or in

case of establishment in the alpine tundra, before a treeline has advanced upslope,

restrictions during these earliest stages can importantly influence the shape and

dynamic of the resulting treeline (Harsch and Bader 2011). Consequently, the

findings presented in this thesis may offer an explanation of observed treeline

Chapter 5

110

patterns, dynamics, and their local variation in the context of site-, life-stage and

species-specific factors. Moreover, they highlight the concurrently causative impact

of such “local” factors in addition to more global drivers such as temperature and the

resulting important implications for the development of predictive models.

OUTLOOK AND FUTURE RESEARCH NEEDS

Observational studies and simple experiments, i.e. comprising only one or few

variables of interest, are the essential foundation and a logical starting point in the

attempt to understand and then be able to predict a natural phenomenon.

Subsequently, the complexity and variability encountered in real landscapes

demands for more integrative approaches to gain a deeper mechanistic insight. In the

debate about treeline formation, this need is reflected in a recent increase of such

integrative studies on treeline tree regeneration as an important driver of treeline

formation. For example, tree-establishment responses are compared at different

temporal scales and life-stages (Bansal and Germino 2009), in multiple species

(Zurbriggen et al. 2013) or to more than one experimentally manipulated

microclimatic factor (Moyes et al. 2013) and combinations of climatic and non-

climatic factors (Grau et al. 2012).

In this spirit, the first two studies of this thesis (Chapter 2 and 3) were

conceived to combine several of those important aspects acting as abiotic and biotic

local factors on treelines in terms of tree regeneration. Thereby, we could present

new insights on the complexity of early establishment responses of treeline trees such

as the idiosyncratic effects of interacting microclimate variables or seasonal changes

of the interaction between tree seedlings and herbaceous alpine vegetation.

Synthesis

111

In the next step, it would be desirable to combine the experimental

manipulation of microclimatic factors and the structure of vegetation cover in a

single field experiment, since it can be expected that the response of both tree

seedlings and alpine vegetation to varying intensities of e.g. temperature and

moisture feeds back to their interaction (positive or negative) and its importance

(Blois et al. 2013). Establishing such an experiment at a site with more stressful

conditions than in our studies (e.g. more summer frost events at higher elevation or

stronger temperature fluctuations and solar radiation loads at tropical alpine

treelines) would allow for longer gradients of abiotic factors, thereby more likely

reaching the tolerance limits of more resistant species (such as P. abies in our studies)

and more appropriately testing hypotheses of plant-plant-interactions such as the

“stress-gradient-hypothesis” with regard to treeline tree seedlings (Bertness and

Callaway 1994). An additional and urgently needed feature to include is an

experimental duration of several years, starting with the initial establishment after

germination and employing a high temporal resolution of measurements. This

would allow to clarify the long-term impact of the seasonal facilitation effects of

alpine herbaceous vegetation we documented and to distinguish between

permanently and irregularly affecting environmental conditions. Finally, it is of

unchanged importance to include different life-stages and tree species to account for

varying treeline regeneration responses caused by such intrinsic factors.

While a project of these dimensions will arguably be difficult to implement

due to the required work load alone, few or no infrastructure and the difficult

accessibility of many treeline sites will further complicate an intended realization.

However, there is a great potential in collaborative initiatives including different

Chapter 5

112

research groups and institutions as known from long-term biodiversity experiments

such as the “Jena-Experiment” (Roscher et al. 2004), which offer important

advantages to large-scale experiments in terms of workforce, temporal stability and

financial aspects. Such an approach should therefore, ideally in the vicinity of an

associated alpine research station (e.g. Lautaret Pass, France; Obergurgl, Austria), be

considered also in the field of treeline research, especially with regard to the

increasing interest in this conspicuous vegetation boundary in the light of climate

change.

A first initiative in this direction was developed with the GTREE-network

(Global Treeline Range Expansion Experiment, Brown et al. 2013), an ongoing joint

effort of different treeline research-groups all over the world, in which we also

participate with a site at a natural treeline close to the Alpine Research Station Joseph

Fourier, Lautaret Pass, France. The aim of GTREE is to disentangle the impact of two

important precursors of treeline range expansion, namely seed availability and

suitability of substrate for establishment and survival, by implementing comparable

seeding experiments at already established treeline research sites worldwide. While

each participant is only responsible for a modest data collection at relatively low

costs, the aggregated, global dataset will allow a wide latitudinal and geographical

comparison of treeline tree regeneration responses. This approach not only allows to

investigate the generality of regeneration limitations on a global scale, but is

subsequently also intended to focus on long-term recruitment patterns and to

perform more region-specific analyses. Thus, studies of this scope could bring

unparalleled contributions for the understanding of the interdependencies that limit

treeline tree regeneration as an important driver of treeline formation.

Synthesis

113

Sometimes the difficulty of conceiving more integrative studies in order to test

hypotheses of treeline formation is not of organizational but of methodological

nature. A promising experimental design evaluated by us in a trial and used in a

study by Fajardo and Piper (2014) aimed at investigating the effect of atmospheric

coupling, i.e. tree-canopy temperature conditions, and vegetation microclimate on

tree growth by growing tree seedlings of different species in suspended or vertically

adjustable pots. The advantage of this approach is that low-temperature limitations

of tree growth ("growth limitation hypothesis", Körner 1998) can be studied in small,

easily replicable units of tree tissue independent from ontogenetic or plant-size

effects and at different heights within and above the alpine vegetation layer.

However, the occurrence of important confounding factors, i.e. high soil temperature

fluctuations and potential moisture deficits due to the comparatively small soil

volume (in spite of reflective outer material, insulation and ventilation layers), does

not allow an unbiased interpretation of results obtained from such a design (Chapter

4). Especially a lack of knowledge concerning the effect of variable versus constant

temperatures on plants in general and tree seedlings in particular conceals the

potential impact of such an artifact, thus urgently calling for further studies

investigating this question. While the presented experimental design may have been

proven unsuitable for the task at hand, alternatives such as the use of active cooling

should be pursued to study the important implications of a low-temperature growth

limit for treeline formation.

114

REFERENCES

Akhalkatsi M, Abdaladze O, Nakhutsrishvili G, Smith WK. 2006. Facilitation of

seedling microsites by Rhododendron caucaskum extends the Betula litwinowii Alpine

treeline, Caucasus Mountains, Republic of Georgia. Arctic Antarctic and Alpine

Research 38: 481–488.

Axelsson PE, Lundmark T, Högberg P, Nordin A. 2014. Belowground competition

directs spatial patterns of seedling growth in boreal pine forests in fennoscandia.

Forests 5: 2106–2121.

Bader MY. 2007. Tropical alpine treelines: how ecological processes control vegetation

patterning and dynamics. PhD Thesis, Wageningen University, Wageningen, The

Netherlands.

Bader MY, Van Geloof I, Rietkerk M. 2007. High solar radiation hinders tree

regeneration above the alpine treeline in northern Ecuador. Plant Ecology 191: 33–45.

Ball MC, Egerton JJG, Leuning R, Cunningham RB. 1997. Microclimate above grass

adversely affects spring growth of seedling snow gum (Eucalyptus pauciflora). Plant,

Cell and Environment 20: 155–166.

Ball MC, Hodges VS, Laughlin GP. 1991. Cold-induced photoinhibition limits

regeneration of snow gum at a tree-line. Functional Ecology 5: 663–668.

Bansal S, Germino MJ. 2009. Temporal variation of nonstructural carbohydrates in

montane conifers: similarities and differences among developmental stages, species

and environmental conditions. Tree physiology 29: 559–568.

Bansal S, Reinhardt K, Germino MJ. 2011. Linking carbon balance to establishment

patterns: Comparison of whitebark pine and Engelmann spruce seedlings along an

herb cover exposure gradient at treeline. Plant Ecology 212: 219–228.

Barbeito I, Dawes MA, Rixen C, Senn J, Bebi P. 2012. Factors driving mortality and

growth at treeline: a 30-year experiment of 92 000 conifers. Ecology 93: 389–401.

115

Barber VA, Juday GP, Finney BP. 2000. Reduced growth of Alaskan white spruce in

the twentieth century from temperature-induced drought stress. Nature 405: 668–673.

Barclay AM, Crawford RMM. 1984. Seedling emergence in the rowan (Sorbus

aucuparia) from an altitudinal gradient. Journal of Ecology 72: 627–636.

Barton AM. 1993. Factors controlling plant distributions: drought, competition, and

fire in montane pines in Arizona. Ecological Monographs 63: 367–397.

Baskin CC, Baskin JM. 2001. Seeds: Ecology, biogeography, and evolution of dormancy

and germination. San Diego: Academic Press.

Batllori E, Camarero JJ. 2009. Seedling recruitment, survival and facilitation in alpine

Pinus uncinata tree line ecotones. Implications and potential responses to climate

warming. Global Ecology and Biogeography 18: 460–472.

Batllori E, Camarero JJ, Gutiérrez E. 2010. Current regeneration patterns at the tree

line in the Pyrenees indicate similar recruitment processes irrespective of the past

disturbance regime. Journal of Biogeography 37: 1938–1950.

Bendix J, Rafiqpoor MD. 2001. Studies on the thermal conditions of soils at the

upper tree line in the Paramo of Papallacta (Estern Cordillera of Ecuador). Erdkunde

55: 257–276.

Bertness MD, Callaway R. 1994. Positive interactions in communities. Trends in

Ecology and Evolution 9: 187–191.

Blois JL, Zarnetske PL, Fitzpatrick MC, Finnegan S. 2013. Climate change and the

past, present, and future of biotic interactions. Science 341: 499–504.

Bloor JMG, Leadley PW, Barthes L. 2008. Responses of Fraxinus excelsior seedlings to

grass-induced above- and below-ground competition. Plant Ecology 194: 293–304.

Bolli JC, Rigling A, Bugmann H. 2007. The influence of changes in climate and land-

use on regeneration dynamics of Norway spruce at the treeline in the Swiss Alps.

Silva Fennica 41: 55–70.

Brändli U-B. 1998. Die häufigsten Waldbäume der Schweiz - Ergebnisse aus dem

116

Landesforstinventar 1983-85: Verbreitung, Standort und Haufigkeit von 30 Baumarten.

Birmensdorf, Switzerland: Eidgenössische Forschungsanstalt tu r Wald, Schnee und

Landschaft.

Brockmann-Jerosch H. 1919. Baumgrenze und Klimacharakter. Schweizerische

Naturforschende Gesellschaft. Pflanzengeographische Kommission. Beiträge zur

geobotanischen Landesaufnahme der Schweiz. Zürich: Rascher.

Brown CD, Johnstone JF, Mamet SD, Trant AJ. 2013. Global Treeline Range

Expansion Experiment field protocols. www.treelineresearch.com.

Brown CD, Vellend M. 2014. Non-climatic constraints on upper elevational plant

range expansion under climate change. Proceedings of the Royal Society-Biological

Sciences 281: 20141779.

Cairns DM, Moen J. 2004. Herbivory influences tree lines. Journal of Ecology 92: 1019–

1024.

Callaway RM, Brooker RW, Choler P, Kidvidze Z, Lortiek CJ, Michalet R, Paolini

L, Pugnaire FI, Newingham B, Aschehoug ET, Armasq C, Kikodze D, Cook BJ.

2002. Positive interactions among alpine plants increase with stress. Nature 417: 844–

848.

Carlsson BA, Callaghan T V. 1991. Positive plant interactions in tundra vegetation

and the importance of shelter. Journal of Ecology, 79: 973–983.

Case BS, Duncan RP. 2014. A novel framework for disentangling the scale-

dependent influences of abiotic factors on alpine treeline position. Ecography 37: 838–

851.

Casper BB, Jackson RB. 1997. Plant competition underground. Annual Review of

Ecology and Systematics 28: 545–570.

Castanha C, Torn MS, Germino MJ, Weibel B, Kueppers LM. 2012. Conifer

seedling recruitment across a gradient from forest to alpine tundra: effects of species,

provenance, and site. Plant Ecology & Diversity 6: 1–12.

117

Cerasoli S, Wertin T, McGuire MA, Rodrigues A, Aubrey DP, Pereira JS, Teskey

RO. 2014. Poplar saplings exposed to recurring temperature shifts of different

amplitude exhibit differences in leaf gas exchange and growth despite equal mean

temperature. AoB PLANTS 6.

Chen I-C, Hill JK, Ohlemüller R, Roy DB, Thomas CD. 2011. Rapid range shifts of

species associated with high levels of climate warming. Science 333: 1024–1026.

Choler P, Michalet R, Callaway RM. 2001. Facilitation and competition on gradients

in alpine plant communities. Ecology 82: 3295–3308.

Cieraad E, McGlone MS, Huntley B. 2014. Southern Hemisphere temperate tree

lines are not climatically depressed. Journal of Biogeography 41: 1456–1466.

Close DC, Beadle CL, Brown PH, Holz GK. 2000. Cold-induced photoinhibition

affects establishment of Eucalyptus nitens (Deane and Maiden) Maiden and Eucalyptus

globulus Labill. Trees - Structure and Function 15: 32–41.

Coomes DA, Grubb PJ. 2000. Impacts of root competition in forests and woodlands:

A theoretical framework and review of experiments. Ecological Monographs 70: 171–

207.

Cuevas JG. 2000. Tree recruitment at the Nothofagus pumilio alpine timberline in

Tierra del Fuego, Chile. Journal of Ecology 88: 840–855.

Cui M, Smith WK. 1991. Photosynthesis, water relations and mortality in Abies-

lasiocarpa seedlings during natural establishment. Tree Physiology 8: 37–46.

Danby RK, Hik DS. 2007. Responses of white spruce (Picea glauca) to experimental

warming at a subarctic alpine treeline. Global Change Biology 13: 437–451.

Daniels LD, Veblen TT. 2004. Spatiotemporal influences of climate on altitudinal

treeline in northern Patagonia. Ecology 85: 1284–1296.

Däniker A. 1923. Biologische Studien über Wald- und Baumgrenze, insbesondere

über die klimatischen Ursachen und deren Zusammenhänge. Vierteljahresschrift der

Naturforschenden Gesellschaft Zürich 68.

118

Daubenmire R. 1954. Alpine timberlines in the Americas and their interpretation.

Butler University Botanical Studies 11: 119–136.

Davis AJ, Jenkinson LS, Lawton JH, Shorrocks B, Wood S. 1998. Making mistakes

when predicting shifts in species range in response to global warming. Nature 391:

783–786.

Dawes MA, Hagedorn F, Handa IT, Streit K, Ekblad A, Rixen C, Körner C,

Hättenschwiler S. 2013. An alpine treeline in a carbon dioxide-rich world: Synthesis

of a nine-year free-air carbon dioxide enrichment study. Oecologia 171: 623–637.

Day ME, Greenwood MS, White AS. 2001. Age-related changes in foliar

morphology and physiology in red spruce and their influence on declining

photosynthetic rates and productivity with tree age. Tree physiology 21: 1195–1204.

Didier L. 2001. Invasion patterns of European larch and Swiss stone pine in

subalpine pastures in the French Alps. Forest Ecology and Management 145: 67–77.

Donohue K, Rubio de Casas R, Burghardt L, Kovach K, Willis CG. 2010.

Germination, postgermination adaptation, and species ecological ranges. Annual

Review of Ecology, Evolution, and Systematics 41: 293–319.

Dufour-Tremblay G, de Vriendt L, Lévesque E, Boudreau S. 2012. The importance

of ecological constraints on the control of multi-species treeline dynamics in Eastern

Nunavik, Québec. American Journal of Botany 99: 1638–1646.

Dullinger S, Dirnböck T, Grabherr G. 2003. Patterns of shrub invasion into high

mountain grasslands of the northern calcareous Alps, Austria. Arctic, Antarctic, and

Alpine Research 35: 434–441.

Ellis RC, Pennington PI. 1992. Factors affecting the growth of Eucalyptus delegatensis

seedlings in inhibitory forest and grassland soils. Plant and Soil 145: 93–105.

Fajardo A, Piper FI. 2014. An experimental approach to explain the Southern andes

elevational treeline. American Journal of Botany 101: 788–795.

Fajardo A, Piper FI, Cavieres LA. 2011. Distinguishing local from global climate

119

influences in the variation of carbon status with altitude in a tree line species. Global

Ecology and Biogeography 20: 307–318.

Fajardo A, Piper FI, Pfund L, Körner C, Hoch G. 2012. Variation of mobile carbon

reserves in trees at the alpine treeline ecotone is under environmental control. New

Phytologist 195: 794–802.

Ferrar P, Cochrane P, Slatyer R. 1988. Factors influencing germination and

establishment of Eucalyptus pauciflora near the alpine tree line. Tree Physiology 4: 27–

43.

Germino MJ, Smith WK. 1999. Sky exposure, crown architecture, and low-

temperature photoinhibition in conifer seedlings at alpine treeline. Plant Cell and

Environment 22: 407–415.

Germino MJ, Smith WK, Resor AC. 2002. Conifer seedling distribution and survival

in an alpine-treeline ecotone. Plant Ecology 162: 157–168.

González de Andrés E, Camarero JJ, Büntgen U. 2015. Complex climate constraints

of upper treeline formation in the Pyrenees. Trees: 941–952.

Grace J. 2002. Impacts of climate change on the tree line. Annals of Botany 90: 537–544.

Grau O, Ninot JM, Blanco-Moreno JM, van Logtestijn RSP, Cornelissen JHC,

Callaghan T V. 2012. Shrub-tree interactions and environmental changes drive

treeline dynamics in the Subarctic. Oikos 121: 1680–1690.

Greenwood S, Chen J-C, Chen C-T, Jump AS. 2015. Temperature and sheltering

determine patterns of seedling establishment in an advancing subtropical treeline.

Journal of Vegetation Science 26: 711–721.

Grubb PJ. 1977. The maintenance of species-richness in plant communities: the

importance of the regeneration niche. Biological Reviews 52: 107–145.

Harper JL. 1977. Population biology of plants. London; New York: Academic Press.

Harsch MA, Bader MY. 2011. Treeline form - a potential key to understanding

treeline dynamics. Global Ecology and Biogeography 20: 582–596.

120

Harsch MA, Hulme PE, McGlone MS, Duncan RP. 2009. Are treelines advancing? A

global meta-analysis of treeline response to climate warming. Ecology Letters 12: 1040–

1049.

Hasselquist N, Germino MJ, McGonigle T, Smith WK. 2005. Variability of

Cenococcum colonization and its ecophysiological significance for young conifers at

alpine-treeline. New Phytologist 165: 867–873.

Hättenschwiler S, Körner C. 1995. Responses to recent climatewarming of Pinus-

sylvestris and Pinus-cembra within their montane transition zone in the Swiss Alps.


He Q, Bertness MD, Altieri AH. 2013. Global shifts towards positive species

interactions with increasing environmental stress. Ecology Letters 16: 695–706.

Heenan PB, Smissen RD. 2013. Revised circumscription of Nothofagus and

recognition of the segregate genera Fuscospora, Lophozonia, and Trisyngyne

(Nothofagaceae). Phytotaxa 146: 1–31.

Hermes K. 1955. Die Lage der oberen Waldgrenze in den Gebirgen der Erde und ihr

Abstand zur Schneegrenze. Kölner Geographische Arbeiten, Heft 5, Geographisches

Institut, Universität Köln.

Hobbie SE, Shevtsova A, Chapin FS. 1999. Plant responses to species removal and

experimental warming in Alaskan tussock tundra. Oikos 84: 417–434.

Hoch G, Körner C. 2003. The carbon charging of pines at the climatic treeline: A

global comparison. Oecologia 135: 10–21.

Hoch G, Körner C. 2009. Growth and carbon relations of tree line forming conifers at

constant vs. variable low temperatures. Journal of Ecology 97: 57–66.

Hoch G, Körner C. 2012. Global patterns of mobile carbon stores in trees at the high-

elevation tree line. Global Ecology and Biogeography 21: 861–871.

Holtmeier F-K. 1965. Die Waldgrenze im Oberengadin in ihrer physiognomischen und

ökologischen Differenzierung. Dissertation, Rheinische Friedrich-Wilhelms-Universität

121

Bonn, Deutschland.

Holtmeier F-K. 1974. Geoökologische Beobachtungen und Studien an der subarktischen und

alpinen Waldgrenze in vergleichender Sicht (nördliches Fennoskandien/Zentralalpen).

Erdwissenschaftliche Forschung, Bd. 8. Wiesbaden.

Holtmeier FK. 2009. Mountain timberlines - ecology, patchiness, and dynamics.

Dordrecht: Springer Netherlands.

Holtmeier FK, Broll G. 2005. Sensitivity and response of northern hemisphere

altitudinal and polar treelines to environmental change at landscape and local scales.

Global Ecology and Biogeography 14: 395–410.

Hoyle GL, Venn SE, Steadman KJ, Good RB, McAuliffe EJ, Williams ER, Nicotra

AB. 2013. Soil warming increases plant species richness but decreases germination

from the alpine soil seed bank. Global Change Biology 19: 1549–61.

von Humboldt A, Bonpland A. 1805. Essai sur la géographie des plantes. Paris: Chez

Levrault, Schoell et Compagnie.

Hutchinson GE. 1957. Concluding remarks. Cold Spring Harbor Symposia on

Quantitative Biology 22: 415–427.

IPCC. 2013. Summary for Policymakers. Climate Change 2013: The Physical Science

Basis. Contribution of Working Group I to the Fifth Assessment Report of the

Intergovernmental Panel on Climate Chan. Cambridge, New York: Cambridge

University Press, .

Islam MA, Macdonald SE. 2004. Ecophysiological adaptations of black spruce (Picea

mariana) and tamarack (Larix laricina) seedlings to flooding. Trees - Structure and

Function 18: 35–42.

Johnson DM, McCulloh KA, Reinhardt K. 2011. The earliest stages of tree growth:

development, physiology and impacts of microclimate. Size- and Age-Related Changes

in Tree Structure and Function. Dordrecht: Springer Netherlands, 65–87.

Kajimoto T, Onodera H, Ikeda S, Daimaru H, Seki T. 1998. Seedling establishment

122

of subalpine stone pine (Pinus pumila) by nutcracker (Nucifraga) seed dispersal on Mt.

Yumori northern Japan. Arctic and Alpine Research 30: 408–417.

Kikvidze Z, Khetsuriani L, Kikodze D, Callaway RM. 2006. Seasonal shifts in

competition and facilitation in subalpine plant communities of the central Caucasus.


Klein JP, Moeschberger ML modifications by JY. 2012. KMsurv: Data sets from

Klein and Moeschberger (1997), Survival Analysis. R package version 0.1-5. URL:

http://cran.r-project.org/package=KMsurv

Kobe RK. 1997. Carbohydrate allocation to storage as a basis of interspecific

variation in sapling survivorship and growth. Oikos 80: 226–233.

Körner C. 1998. A re-assessment of high elevation treeline positions and their

explanation. Oecologia 115: 445–459.

Körner C. 2012. Alpine treelines. Functional ecology of the global high elevation tree limits.

Basel: Springer.

Körner C, Hoch G. 2006. A test of treeline theory on a montane permafrost island.

Arctic Antarctic and Alpine Research 38: 113–119.

Körner C, Paulsen J. 2004. A world-wide study of high altitude treeline

temperatures. Journal of Biogeography 31: 713–732.

Kozlowski T. 1992. Carbohydrate sources and sinks in woody plants. Botanical

Review 58: 107–222.

Krebs C. 1972. Ecology. The experimental analysis of distribution and abundance. New

York: Harper & Row.

Kroiss SJ, HilleRisLambers J. 2015. Recruitment limitation of long-lived conifers:

implications for climate change responses. Ecology 96: 1286–1297.

Kullman L. 2007. Tree line population monitoring of Pinus sylvestris in the Swedish

Scandes, 1973-2005: implications for tree line theory and climate change ecology.

Journal of Ecology 95: 41–52.

123

Larigauderie A, Ellis BA, Mills JN, Kummerow J. 1991. The effect of root and shoot

temperatures on growth of Ceanothus greggii seedlings. Annals of Botany 67: 97–101.

LeBarron RK. 1945. Adjustment of black spruce root systems to increasing depth of

peat. Ecological Society of America 26: 309–311.

Lenz A, Hoch G, Körner C. 2013. Early season temperature controls cambial activity

and total tree ring width at the alpine treeline. Plant Ecology & Diversity 6: 365–375.

Leuzinger S, Körner C. 2007. Tree species diversity affects canopy leaf temperatures

in a mature temperate forest. Agricultural and Forest Meteorology 146: 29–37.

Li M, Hoch G, Körner C. 2002. Source/sink removal affects mobile carbohydrates in

Pinus cembra at the Swiss treeline. Trees - Structure and Function 16: 331–337.

Li Q, Song Y, Li G, Yu P, Wang P, Zhou D. 2015. Grass-legume mixtures impact soil

N, species recruitment, and productivity in temperate steppe grassland. Plant and Soil

394: 271–285.

Lloyd AH, Bunn AG. 2007. Responses of the circumpolar boreal forest to 20th

century climate variability. Environmental Research Letters 2: 045013 (13pp).

Løken A. 1959. [Germination experiments in refrigerator chamber]. Meddedels. f.

Vestlandets forstl. Forsøksstasjon 11: 1–19.

Maher EL, Germino MJ. 2006. Microsite differentiation among conifer species during

seedling establishment at alpine treeline. Ecoscience 13: 334–341.

Maher EL, Germino MJ, Hasselquist NJ. 2005. Interactive effects of tree and herb

cover on survivorship, physiology, and microclimate of conifer seedlings at the

alpine tree-line ecotone. Canadian Journal of Forest Research 35: 567–574.

Malanson GP, Brown DG, Butler DR, Cairns DM, Fagre DB, Walsh SJ. 2009.

Ecotone dynamics: invasibility of alpine tundra by tree species from the subalpine

forest. The changing alpine treeline — The example of Glacier National Park, MT, USA.

Developments in earth surface processes. Vol. 12. Amsterdam: Elsevier, 35–57.

Malanson GP, Resler LM, Bader MY, Holtmeier FK, Butler DR, Weiss DJ, Daniels

124

LD, Fagre DB. 2011. Mountain Treelines: a Roadmap for Research Orientation. Arctic

Antarctic and Alpine Research 43: 167–177.

Marek R. 1910. Waldgrenzstudien in den österreichischen Alpen. Petermanns

geographische Mitteilungen, Ergänzungsband 36, Heft 168. Gotha: Justus Perthes.

Marion G, Henry G, Freckman D, Johnstone J, Jones G, Jones M, Lévesque E,

Molau U, Molgaard P, Parsons A, Svoboda J, Virginia R. 1997. Open-top designs for

manipulating field temperature in high-latitude ecosystems. Global Change Biology 3:

20–32.

Mark AF, Porter S, Piggott JJ, Michel P, Maeglp T, Dickinson KJM. 2008.

Altitudinal patterns of vegetation, flora, life forms, and environments in the alpine

zone of the Fiord Ecological Region, New Zealand. New Zealand Journal of Botany 46:

205–237.

Mattes H. 1994. Coevolutional aspects of stone pines and nutcrackers. International

workshop on subalpine stone pines and their environment: the status of our knowledge. INT-

GTR-309. St. Moritz, Switzerland, September 5-11,1992. UT, USA: Intermountain

Research Station, 31–35.

Matyssek R. 1986. Carbon, water and nitrogen relations in evergreen and deciduous

conifers. Tree physiology 2: 177–187.

McNair JN, Sunkara A, Frobish D. 2012. How to analyse seed germination data

using statistical time-to-event analysis: non-parametric and semi-parametric

methods. Seed Science Research 22: 77–95.

Melillo JM, Steudler P a, Aber JD, Newkirk K, Lux H, Bowles FP, Catricala C,

Magill A, Ahrens T, Morrisseau S. 2002. Soil warming and carbon-cycle feedbacks

to the climate system. Science 298: 2173–2176.

Michaelis P. 1934. Ökologische Studien an der alpinen Baumgrenze V. Osmotischer

Wert und Wassergehalt während des Winters in verschiedenen Höhenlagen. Jahrbuch

für wissenschaftliche Botanik 80: 337–362.

Michalet R, Le Bagousse-Pinguet Y, Maalouf JP, Lortie CJ. 2014. Two alternatives to

125

the stress-gradient hypothesis at the edge of life: The collapse of facilitation and the

switch from facilitation to competition. Journal of Vegetation Science 25: 609–613.

Moir WH, Rochelle SG, Schoettle AW. 1999. Microscale patterns of tree

establishment near upper treeline, Snowy Range, Wyoming, U.S.A. Arctic and Alpine

Research 31: 379–388.

Molina-Montenegro MA, Gallardo-Cerda J, Flores TSM, Atala C. 2012. The trade-

off between cold resistance and growth determines the Nothofagus pumilio treeline.

Plant Ecology 213: 133–142.

Moyes AB, Castanha C, Germino MJ, Kueppers LM. 2013. Warming and the

dependence of limber pine (Pinus flexilis) establishment on summer soil moisture

within and above its current elevation range. Oecologia 171: 271–282.

Nambiar EKS. 1990. Interplay between nutrients, water, root growth and

productivity in young plantations. Forest Ecology and Management 30: 213–232.

Noble DL, Alexander RR. 1977. Environmental factors affecting natural regeneration

of Engelmann spruce in the central Rocky Mountains. Forest Science 23: 420–429.

Ohse B, Jansen F, Wilmking M. 2012. Do limiting factors at Alaskan treelines shift

with climatic regimes? Environmental Research Letters 7: 015505.

Ozenda P. 1988. Die Vegetation der Alpen im europäischen Gebirgsraum. Stuttgart; New

York: Gustav Fischer Verlag.

Palacio S, Hoch G, Sala A, Körner C, Millard P. 2014. Does carbon storage limit tree

growth? New Phytologist 201: 1096–1100.

Parmesan C, Yohe G. 2003. A globally coherent fingerprint of climate change

impacts across natural systems. Nature 421: 37–42.

Paulsen J, Körner C. 2014. A climate-based model to predict potential treeline

position around the globe. Alpine Botany: 1–12.

Perry KB. 1998. Basics of frost and freeze protection for horticultural crops.

HortTechnology 8: 10–15.

126

Petit G, Anfodillo T, Carraro V, Grani F, Carrer M. 2011. Hydraulic constraints limit

height growth in trees at high altitude. New Phytologist 189: 241–252.

Piper FI, Cavieres L a., Reyes-Díaz M, Corcuera LJ. 2006. Carbon sink limitation and

frost tolerance control performance of the tree Kageneckia angustifolia D. Don

(Rosaceae) at the treeline in central Chile. Plant Ecology 185: 29–39.

Piper FI, Fajardo A, Cavieres LA. 2013. Simulated warming does not impair seedling

survival and growth of Nothofagus pumilio in the southern Andes. Perspectives in Plant

Ecology, Evolution and Systematics 15: 97–105.

Pisek A, Cartellieri E. 1939. Zur Kenntnis des Wasserhaushaltes der Pflanzen, IV.

Bäume und Sträucher. Jahrbuch für wissenschaftliche Botanik 88: 22–68.

Popp M, Lied W, Meyer AJ, Richter A, Schiller P, Schwitte H. 1996. Sample

preservation for determination of organic compounds: microwave versus freeze-

drying. Journal of Experimental Botany, 47: 1469–1473.

Pregitzer KS, King JS, Burton AJ, Brown SE. 2000. Responses of tree fine roots to

temperature. New Phytologist 147: 105–115.

Qi Z, Liu H, Wu X, Hao Q. 2015. Climate-driven speedup of alpine treeline forest

growth in the Tianshan Mountains, Northwestern China. Global Change Biology 21:

816–826.

Quentin AG, Pinkard EA, Ryan MG, Tissue DT, Baggett LS, Adams HD, Maillard

P, Marchand J, Landhäusser SM, Lacointe A, Gibon Y, Anderegg WRL, Asao S,

Atkin OK, Bonhomme M, Claye C, Chow PS, Clément-Vidal A, Davies NW,

Dickman LT, Dumbur R, Ellsworth DS, Falk K, Galiano L, Grünzweig JM,

Hartmann H, Hoch G, Hood S, Jones JE, Koike T, Kuhlmann I, Lloret F, Maestro

M, Mansfield SD, Martínez-Vilalta J, Maucourt M, McDowell NG, Moing A,

Muller B, Nebauer SG, Niinemets Ü, Palacio S, Piper F, Raveh E, Richter A,

Rolland G, Rosas T, Saint Joanis B, Sala A, Smith RA, Sterck F, Stinziano JR,

Tobias M, Unda F, Watanabe M, Way DA, Weerasinghe LK, Wild B, Wiley E,

Woodruff DR. 2015. Non-structural carbohydrates in woody plants compared

among laboratories. Tree Physiology 35: 1146–1165.

127

Quétier F, Lavorel S, Thuiller W, Davies I. 2007. Plant-trait-based modeling

assessment of ecosystem-service sensitivity to land-use change. Ecological Applications

17: 2377–2386.

R Core Team. 2015. R: A Language and Environment for Statistical Computing. R

version 3.2.2. URL: http://www.r-project.org/

Rameau J, Dumé G, Mansion D. 1993. Flore forestie re francaise: guide e cologique illustre.

2. Montagnes. Paris: Institut pour le Développement Forestier.

Raspé O, Findlay C, Jacquemart AL. 2000. Sorbus aucuparia L. Journal of Ecology 88:

910–930.

Rolland C, Petitcolas V, Michalet R. 1998. Changes in radial tree growth for Picea

abies, Larix decidua, Pinus cembra and Pinus uncinata near the alpine timberline since

1750. Trees 13: 40–53.

Roscher C, Schumacher J, Baade J. 2004. The role of biodiversity for element cycling

and trophic interactions: an experimental approach in a grassland community. Basic

and Applied Ecology 121: 107–121.

Rossi S, Deslauriers A, Griçar J, Seo JW, Rathgeber CBK, Anfodillo T, Morin H,

Levanic T, Oven P, Jalkanen R. 2008. Critical temperatures for xylogenesis in

conifers of cold climates. Global Ecology and Biogeography 17: 696–707.

Ryser P. 1993. Influences of neighbouring plants on seedling establishment in

limestone grassland. Journal of Vegetation Science 4: 195–202.

Schupp EW. 1995. Seed-seedling conflicts, habitat choice, and patterns of plant

recruitment. American Journal of Botany 82: 399–409.

Scott P, Bentley C, Fayle D, Hansell R. 1987. Crown forms and shoot elongation of

white spruce at the treeline, Churchill, Manitoba, Canada. Arctic and Alpine Research

19: 175–186.

Seiwa K. 1998. Advantages of early germination for growth and survival of seedlings

of Acer mono under different overstorey phenologies in deciduous broad-leaved

128

forests. Journal of Ecology 86: 219–228.

Seiwa K, Miwa Y, Sahashi N, Kanno H, Tomita M, Ueno N, Yamazaki M. 2008.

Pathogen attack and spatial patterns of juvenile mortality and growth in a temperate

tree, Prunus grayana. Canadian Journal of Forest Research 38: 2445–2454.

Senn J. 1999. Tree mortality caused by Gremmeniella abietina in a subalpine

afforestation in the central Alps and its relationship with duration of snow cover.

European Journal of Forest Pathology 29: 65–74.

Shen W, Zhang L, Liu X, Luo T. 2014. Seed-based treeline seedlings are vulnerable to

freezing events in the early growing season under a warmer climate: Evidence from a

reciprocal transplant experiment in the Sergyemla Mountains, southeast Tibet.

Agricultural and Forest Meteorology 187: 83–92.

Shi P, Körner C, Hoch G. 2008. A test of the growth-limitation theory for alpine tree

line formation in evergreen and deciduous taxa of the eastern Himalayas. Functional

Ecology 22: 213–220.

Shiyatov SG, Terent’ev MM, Fomin V V., Zimmermann NE. 2007. Altitudinal and

horizontal shifts of the upper boundaries of open and closed forests in the polar

Urals. Russian Journal of Ecology 38: 223–227.

Sims HP, Mueller-Dombois D. 1968. Effect of grass competition and depth to water

table on height growth of coniferous tree seedlings. Ecology 49: 597–603.

Smith WK, Germino MJ, Hancock TE, Johnson DM. 2003. Another perspective on

altitudinal limits of alpine timberlines. Tree Physiology 23: 1101–1112.

Steiner M. 1935. Winterliches Bioklima und Wasserhaushalt an der alpinen

Waldgrenze. Bioklimatische Beiblätter 2: 57–65.

Stevens GC, Fox JF. 1991. The causes of treeline. Annual Review of Ecology and

Systematics 22: 177–191.

Strand M, Lofvenius M, Bergsten U, Lundmark T, Rosvall O. 2006. Height growth

of planted conifer seedlings in relation to solar radiation and position in Scots pine

129

shelterwood. Forest Ecology and Management 224: 258–265.

Sullivan PF, Sveinbjörnsson B. 2010. Microtopographic control of treeline advance

in Noatak National Preserve, Northwest Alaska. Ecosystems 13: 275–285.

Sutherst RW, Maywald GF, Bourne AS. 2007. Including species interactions in risk

assessments for global change. Global Change Biology 13: 1843–1859.

Tegischer K, Tausz M, Wieser G, Grill D. 2002. Tree- and needle-age-dependent

variations in antioxidants and photoprotective pigments in Norway spruce needles

at the alpine timberline. Tree Physiology 22: 591–596.

Thébault A, Clément JC, Ibanez S, Roy J, Geremia RA, Pérez CA, Buttler A,

Estienne Y, Lavorel S. 2014. Nitrogen limitation and microbial diversity at the

treeline. Oikos 123: 729–740.

Therneau T. 2015. A Package for Survival Analysis in S. version 2.38. R package

version 2.38-1. URL: http://cran.r-project.org/package=survival

Tranquillini W. 1979. Physiological ecology of the alpine timberline. Berlin, Heidelberg:

Springer.

Troll C. 1973. The upper timberlines in different climatic zones. Arctic and Alpine

Research 5: 3–18.

Ulber M, Gugerli F, Bozic G. 2004. EUFORGEN Technical Guidelines for genetic

conservation and use for Swiss stone pine (Pinus cembra). Rome: International Plant

Genetic Resources Institute.

Venn SE, Morgan JW, Green PT. 2009. Do facilitative interactions with neighboring

plants assist the growth of seedlings at high altitudes in alpine Australia? Arctic,

Antarctic, and Alpine Research 41: 381–387.

Vittoz P, Rulence B, Largey T, Freléchoux F. 2008. Effects of climate and land-use

change on the establishment and growth of Cembran pine (Pinus cembra L.) over the

altitudinal treeline ecotone in the central Swiss Alps. Arctic, Antarctic, and Alpine

Research 40: 225–232.

130

Wahlenberg G. 1813. De vegetatione et climate in Helvetia septentrionali inter flumina

Rhenum et Arolam observatis et cum summi septentrionis comparatis tentamen. Zürich:

impensis Orell, Füssli et socc.

Wang Y, Zwiazek JJ. 1999. Effects of early spring photosynthesis on carbohydrate

content, bud flushing and root and shoot growth of Picea glauca bareroot seedlings.

Scandinavian Journal of Forest Research 14: 295–302.

Wardle P. 1974. Alpine timberlines. Arctic and alpine environments. London: Methuen,

371–402.

Wardle P. 1985. New Zealand timberlines. 1. Growth and survival of native and

introduced tree species in the Craigiebum Range, Canterbury. New Zealand Journal of

Botany 23: 219–234.

Wardle P. 1993. Causes of alpine timberline: A review of the hypotheses. Forest

development in cold climates. New York, USA: Plenum Press, 89–103.

Wardle P. 2008. New Zealand forest to alpine transitions in global context. Arctic

Antarctic and Alpine Research 40: 240–249.

Weih M, Karlsson PS. 2001. Growth response of Mountain birch to air and soil

temperature: Is increasing leaf-nitrogen content an acclimation to lower air

temperature? New Phytologist 150: 147–155.

Wick L, Tinner W. 1997. Vegetation changes and timberline fluctuations in the

central Alps as indicators of Holocene climatic oscillations. Arctic and Alpine Research

29: 445–458.

Wieser G, Tausz M. 2007. Trees at their upper limit - Treelife limitation at the alpine

timberline. Dordrecht: Springer Netherlands.

Wiley E, Helliker B. 2012. A re-evaluation of carbon storage in trees lends greater

support for carbon limitation to growth. New Phytologist 195: 285–289.

Wisz MS, Pottier J, Kissling WD, Pellissier L, Lenoir J, Damgaard CF, Dormann

CF, Forchhammer MC, Grytnes JA, Guisan A, Heikkinen RK, Høye TT, Kühn I,

131

Luoto M, Maiorano L, Nilsson MC, Normand S, Öckinger E, Schmidt NM,

Termansen M, Timmermann A, Wardle DA, Aastrup P, Svenning JC. 2013. The role

of biotic interactions in shaping distributions and realised assemblages of species:

implications for species distribution modelling. Biological Reviews 88: 15–30.

Xu Z, Hu T, Zhang Y. 2012. Effects of experimental warming on phenology, growth

and gas exchange of treeline birch (Betula utilis) saplings, Eastern Tibetan Plateau,

China. European Jounal of Forest Research 131: 811–819.

Yamazaki M, Iwamoto S, Seiwa K. 2009. Distance- and density-dependent seedling

mortality caused by several diseases in eight tree species co-occurring in a temperate

forest. Plant Ecology 201: 181–196.

Zurbriggen N, Hättenschwiler S, Frei ES, Hagedorn F, Bebi P. 2013. Performance of

germinating tree seedlings below and above treeline in the Swiss Alps. Plant Ecology

214: 385–396.

Zuur AF, Ieno EN, Walker NJ, Saveliev AA, Smith GM. 2009. Mixed Effects Models

and Extensions in Ecology with R. New York: Springer.

Zywiec M, Holeksa J, Wesolowska M, Szewczyk J, Zwijacz-Kozica T, Kapusta P.

2013. Sorbus aucuparia regeneration in a coarse-grained spruce forest - a landscape

scale. Journal of Vegetation Science 24: 735–743.

Zywiec M, Ledwoń M. 2008. Spatial and temporal patterns of rowan (Sorbus

aucuparia L.) regeneration in West Carpathian subalpine spruce forest. Plant Ecology

194: 283–291.

132

SUMMARY

Alpine treelines are important and conspicuous vegetation boundaries on high

mountains, which have caught the interest of naturalists and researchers for many

centuries. They are ultimately limited by heat deficiency and thus expected to

advance with the currently warming climate. However, treeline responses are locally

contrasting due to other climatic factors, tree species ecology and life-stage-

dependent responses. Since regeneration of trees is a prerequisite for treelines to

remain stable or to move upslope, especially the critical earliest life-stages of

germination and seedling establishment may present a major life-history bottleneck

for treeline tree populations. Studying the limitations of tree recruitment is therefore

an important step to gain a more mechanistic understanding of treeline formation

and to reliably predict their future dynamics.

The first two chapters of this thesis present studies investigating the effect of abiotic

and biotic environmental factors, respectively, on the two critical regeneration stages

of germination and early seedling establishment in five important European treeline

tree species. They show that local environmental factors in terms of microclimate and

biotic interactions exert an important impact on these earliest life-stages: i) early tree

establishment was either limited by temperature or moisture and often by

interactions of both, and ii) negative effects in interaction with herbaceous alpine

vegetation dominated, but evergreen tree species could benefit from a seasonal

switch to facilitation in autumn. While the responses to microclimate were generally

consistent over both early life-stages, only seedlings were affected by interactions

with herbaceous vegetation, and overall idiosyncratic patterns were predominant.

133

The important and variable impact of these site-specific and intrinsic factors on

treeline tree regeneration may further help to explain observed treeline patterns and

highlights the importance of their consideration for predictive models.

The third chapter presents a commentary contesting a recent study with a new,

promising approach of testing the low-temperature growth limitation of treeline

trees by growing tree seedlings in suspended pots. The criticism is based on the own

previous identification of considerable confounding factors in a similar experimental

design and the accordingly questionable interpretation of the study’s results is re-

evaluated with regard to the impact of high soil temperature fluctuations and

moisture deficiency. Additionally, species-specific treeline-forming mechanisms are

reviewed in relation to the study’s focal species, and alternatives for the after all

elegant method are explored.

The thesis concludes with an outlook in the potential of large-scale collaborative

projects to meet the need of more integrative studies that consider interacting

environmental factors, life-stage- and species-specific limitations as well as seasonal

and long-term responses. Such studies could bring unparalleled contributions for the

understanding of the interdependencies that limit treeline tree regeneration as an

important driver of treeline formation.

134

ZUSAMMENFASSUNG

Alpine Baumgrenzen sind wichtige und auffällige Vegetationsgrenzen hoher Berge,

die seit Jahrhunderten das Interesse von Naturforschern und Wissenschaftlern auf

sich ziehen. Sie sind schlussendlich durch einen Mangel an Wärme limitiert, was eine

Ausdehnung durch die aktuelle Klimaerwärmung erwarten lässt. Dennoch findet

man lokal stark variierende Muster und Dynamiken, welche durch andere

klimatische Faktoren und spezifische Bedürfnisse von Baumart oder Phase des

Lebenszyklus beeinflusst werden. Da die Regeneration von Bäumen die

Vorbedingung ist, damit Baumgrenzen stabil bleiben oder bergan voranschreiten,

könnten vor allem die kritischen Phasen der Keimung und Keimlingsetablierung

einen bedeutenden, lebensgeschichtlichen Engpass für Baumpopulationen der

Baumgrenze darstellen. Die Erforschung der Faktoren, die die Regeneration von

Bäumen limitieren ist daher ein wichtiger Schritt zu einem tieferen mechanistischen

Verständnis der Entstehung von Baumgrenzen und zu zuverlässigen Vorhersagen

ihrer zukünftigen Dynamik.

Die ersten beiden Kapitel dieser Dissertation präsentieren zwei Studien, welche

einmal die Wirkung abiotischer und einmal biotischer Umweltfaktoren auf die

beiden kritischen Regenerationsphasen Keimung und Keimlingsetablierung für fünf

wichtige europäische Baumarten der alpinen Baumgrenze untersuchen. Sie zeigen,

dass lokale Umweltfaktoren in Form von Mikroklima und biotischen Interaktionen

einen starken Einfluss auf die frühen Lebensphasen ausüben: i) frühe

Keimlingsetablierung war entweder durch Temperatur oder Feuchte, und häufig

135

durch Interaktionen beider Faktoren limitiert, und ii) negative Effekte durch die

Interaktion mit alpiner Graslandvegetation überwogen, aber immergrüne Baumarten

konnten im Herbst von einem saisonalen Wechsel zu einer förderlichen Beziehung

profitieren. Während die Reaktionen auf das Mikroklima generell in beiden frühen

Lebensphasen übereinstimmten, wurden nur Keimlinge von Interaktionen mit der

Graslandvegetation beeinträchtigt, und insgesamt überwogen artspezifische Effekte.

Der wichtige und variable Einfluss dieser ortsspezifischen und intrinsischen

Faktoren auf die Regeneration von Bäumen der Baumgrenze kann so im Weiteren

helfen, deren beobachtete Muster zu erklären und hebt hervor, wie wichtig ihre

Berücksichtigung in voraussagenden Modellen ist.

Das dritte Kapitel präsentiert einen Kommentar, der eine rezente Studie in Frage

stellt, in welcher das Anziehen von Keimlingen in hoch aufgehängten Behältnissen

als neuer, vielversprechender Ansatz zum Testen der Kältelimitierung des

Baumwachstums verwendet wird. Die Kritik basiert auf der eigenen vorherigen

Feststellung von beträchtlichen Störfaktoren in einem ähnlichen experimentellen

Aufbau, und die demzufolge fraglichen Interpretationen der Ergebnisse dieser Studie

werden unter der Berücksichtigung des Einflusses von starken

Bodentemperaturschwankungen und Wassermangel neu bewertet. Zusätzlich

werden artspezifische Mechanismen der Baumgrenzenbildung im Hinblick auf die

Zielarten der besagten Studie besprochen und Alternativen für die trotz allem

elegante Methode untersucht.

Die Dissertation schließt mit einem Ausblick auf das Potential von großangelegten,

gemeinschaftlichen Projekten um den Bedarf an integrativeren Studien zu decken,

136

welche interagierende Umweltfaktoren, art- und lebensphasenspezifische

Limitierungen sowie saisonale und Langzeitreaktionen berücksichtigen. Solche

Studien könnten einmalige Beiträge zum Verständnis der Wechselbeziehungen

leisten, welche die Regeneration von Bäumen als wichtigen Einflussfaktor der

Baumgrenzenbildung limitieren.

137

DANKSAGUNG

An dieser Stelle möchte ich mich zunächst bei meinen beiden Betreuern Maaike

Bader und Gerhard Zotz mit viel Nachdruck bedanken, die mir mit diesem

Promotionsprojekt eine sehr erfahrungsreiche und auch schöne Zeit in der

einmaligen Naturlandschaft der französischen Alpen ermöglicht haben, und deren

immer offene Türen / Ohren sowie unkomplizierte Email- / Skype-Kommunikation

ich sehr zu schätzen wusste!

Le meilleur concept ne peut pas aller bien loin sans un endroit à la hauteur de sa

mise en œuvre – cet endroit nous a été gracieusement fourni par la Station Alpine

Joseph Fourier de l’Université de Grenoble avec les zones expérimentales et le

Chalet-Labo au Jardin Alpin du Lautaret. Je remercie tout particulièrement l’ancien

directeur du Jardin Alpin, Serge Aubert, à qui le support de la recherche et la

réalisation de mes idées et de mes manips tenaient toujours très à cœur. Un grand

merci aussi à Karl Grigulis, Franck Delbart et Pascal Salze pour tout support

technique et administratif, et à toute l’équipe du Lautaret pour les bons moments

passés ensemble et les soirées mémorables dans le chalet des jardiniers.

Ein großer Dank gilt auch meinem Vorgänger Marc Müller, von dessen ersten

Erfahrungen ich bei der Planung meiner Experimente sehr profitieren konnte, und

allen helfenden Händen, ohne die ich dieses Projekt nicht hätte realisieren können!

Allen voran meine resoluten Feld-Hiwis Verena Schenk und Eric Thurm, Jasmin

Baruck und Gesa Pries, Carla Sardemann, Mathilde Vicente und Tizian Weichgrebe,

die jeweils über eine Feldsaison zusammen mit mir Wetter und Anstrengungen

138

getrotzt haben um die Experimente in Frankreich aufzubauen, zu pflegen und um

Messungen durchzuführen. Auch in Oldenburg hatte ich viel Unterstützung durch

die Hiwis Sven Wemken, Elif Gökpinar, Nawin Grabowsky, Nora Wissner, unsere

findigen TA’s Ingeborg Eden und Brigitte Rieger, sowie auf die eine oder andere Art

und Weise alle Mitglieder der AG’s Zotz und Albach. Vielen Dank an alle!

Schließlich hat meine Familie einen besonders großen Anteil an den Resultaten der

letzten Jahre:

Meine wunderbaren Eltern Horst und Hiltrud Kern, die schon früh mein Interesse an

der Natur gefördert haben und mich während meiner ganzen Entwicklung,

Ausbildung und auf allen neuen Wegen nach Kräften unterstützt haben, und mein

wunderbarer Bruder Boris, der mir immer wieder einen frischen und

unkonventionellen Blick auf die Dinge ermöglicht.

Et surtout mon merveilleux mari Jessy, qui était à mes côtés à travers toutes les

étapes de cette thèse, du terrain dans les hauteurs des Alpes jusque dans le labo à

Oldenburg, des moments les plus joyeux jusqu’au temps le plus difficile. Merci mon

amour d’avoir tenu ma corde de secours pendant cette escalade exigeante, j’ai

beaucoup hâte à nos aventures à venir!

139

LEBENSLAUF

PERSÖNLICHE DATEN

Name: Hannah Loranger

Geboren: 26.11.1985, Duisburg

Familienstand: verheiratet

Staatsangehörigkeit: Deutsch

WISSENSCHAFTLICHER WERDEGANG

6.2005 Abitur, Gymnasium Thomaeum, Kempen

10.2005 – 8.2007 Grundstudium Biologie an der Universität Osnabrück

9.2007 – 8.2008 Auslandsstudium an der Partneruniversität Université de

Sherbrooke, Québec, Kanada

9.2008 – 6.2011 Hauptstudium Biologie an der Universität Osnabrück mit dem

Abschluss Diplom, Prüfungsfächer: Botanik, Ökologie, Pflanzen-

physiologie, externe Diplomarbeit „Increasing invertebrate

herbivory along an experimental grassland plant diversity

gradient“ an der Universität Jena unter der Leitung von Prof. Dr.

Wolfgang W. Weisser und Jun. Prof. Dr. Till Eggers

7.2011 – 9.2011 Wissenschaftliche Hilfskraft am Institut für Ökologie, Universität

Jena

Seit 10.2011 Promotion an der Carl-von-Ossietzky Universität Oldenburg im

DFG-Projekt „The regeneration niche of trees at the alpine

treeline: climatic constraints on germination and seedling

establishment”

Leitung: Prof. Dr. Gerhard Zotz

140

PUBLIKATIONEN

Loranger J, Meyer ST, Shipley B, Kattge J, Loranger H, Roscher C & Weisser WW.

2012. Predicting invertebrate herbivory from plant traits: evidence from 51

grassland species in experimental monocultures. Ecology 93: 2674-2682.

Loranger J, Meyer ST, Shipley B, Kattge J, Loranger H, Roscher C, Wirth C & Weisser

WW. 2013. Predicting invertebrate herbivory from plant traits: Polycultures show

strong nonadditive effects. Ecology 94: 1499-1509.

Loranger H, Weisser WW, Ebeling A, Eggers T, De Luca E, Loranger J, Roscher C,

Meyer ST. 2014. Invertebrate herbivory increases along an experimental gradient

of grassland plant diversity. Oecologia 174: 183-193.

Bader M, Loranger H & Zotz G. 2014. A cool experimental approach to explain

elevational treelines. 2014. But can it explain them? American Journalof Botany

101 (9): 1403-1408.

KONFERENZBEITRÄGE

Loranger H, Zotz G & Bader 2014. Species-specific climate responses in tree

regeneration at the alpine treeline. Joint annual Meeting British Ecological

Society and Société Française d’Écologie, 9 – 12 December 2014, Lille, France

141

AUTHORS’ CONTRIBUTIONS

In the following section I present the contributions of all authors to the chapters 2 to

4 and their respective manuscripts or publications.

Chapter 2: Loranger H, Zotz G, Bader M Y. Impacts of soil microclimate on early

establishment of trees at the alpine treeline: idiosyncratic responses and the

importance of soil moisture. Resubmitted to AoB Plants

MB conceived the study with input from GZ, HL and MB planned the

experimental design, HL set up the experiment, collected the data, performed the

statistical analysis of the data and produced graphs and tables, HL wrote the initial

version of the manuscript with the support of MB and GZ, all authors contributed to

the revision of the manuscript.

Chapter 3: Loranger H, Zotz G, Bader M Y. Competitor or facilitator? The role of

grassland vegetation for germination and seedling performance of tree species at the

alpine treeline. Submitted to Functional Ecology

MB conceived the study with input from GZ, HL and MB planned the

experimental design, HL set up the experiment, collected the data, performed the

statistical analysis of the data and produced graphs and tables, HL wrote the initial

version of the manuscript with the support of MB and GZ.

Chapter 4: Bader M Y, Loranger H, Zotz, G. A cool experimental approach to explain

elevational treelines, but can it explain them? American Journal of Botany 101(9): 1–6

142

MB, HL and GZ planned the initial study, HL designed a prototype planting-

tube, collected and evaluated preliminary data, HL and MB decided to cancel the

experiment due to confounding factors arising from the specific design, MB wrote a

commentary contesting a study using a similar design without controlling for

confounding factors (Fajardo and Piper 2014, American Journal of Botany, 101: 788–795)

with the participation of GZ and HL.

Als Betreuer der Arbeit bestätige ich die Richtigkeit der Autorenbeiträge zu den

aufgeführten Kapiteln bzw. deren Manuskripten oder Veröffentlichungen.

…………………………………..

Prof. Dr. Gerhard Zotz

Hiermit bestätige ich die Richtigkeit der Autorenbeiträge zu den aufgeführten

Kapiteln bzw. deren Manuskripten oder Veröffentlichungen.

…………………………………..

Hannah Loranger

143

ALLGEMEINE ERKLÄRUNG

Ich füge folgende Erklärungen an gemäß § 11 Abs. 2 der Promotionsordnung (Stand:

13.01.2013) der Fakultät für Mathematik und Naturwissenschaften, Carl von

Ossietzky

Universität Oldenburg:

- Die Dissertation “The regeneration niche of trees at the alpine treeline -

Constraints of microclimate and the alpine grassland vegetation on

germination and seedling establishment” wurde von mir selbständig und nur

unter Verwendung der angegebenen Hilfsmittel verfasst.

- Ich strebe eine Promotion zur Doktorin der Biologie (Dr. rer. nat) an.

- Teile der Dissertation (Kapitel 4) wurden bereits veröffentlicht (siehe Author’s

contributions).

- Die Dissertation lag/liegt weder in ihrer Gesamtheit noch in Teilen einer

anderen wissenschaftlichen Hochschule zur Begutachtung in einem

Promotionsverfahren vor.

Hiermit bestätige ich die Richtigkeit der allgemeinen Erklärung.

…………………………………………

Hannah Loranger

I plant the first seedlings into the experiment, Lautaret June 2012

Documents

CONSTRAINTS OF MICROCLIMATE AND THE ALPINE … · stressful conditions for the focal organisms. Alpine treelines are among the most conspicuous vegetation limits in existence, and