UNIVERSITY OF OULU P .O. Box 8000 F I -90014 UNIVERSITY OF OULU FINLAND

A C T A U N I V E R S I T A T I S O U L U E N S I S

University Lecturer Tuomo Glumoff

University Lecturer Santeri Palviainen

Postdoctoral research fellow Sanna Taskila

Professor Olli Vuolteenaho

University Lecturer Veli-Matti Ulvinen

Planning Director Pertti Tikkanen

Professor Jari Juga

University Lecturer Anu Soikkeli

Professor Olli Vuolteenaho

Publications Editor Kirsti Nurkkala

ISBN 978-952-62-1509-9 (Paperback)ISBN 978-952-62-1510-5 (PDF)ISSN 0355-3191 (Print)ISSN 1796-220X (Online)

U N I V E R S I TAT I S O U L U E N S I SACTAA

SCIENTIAE RERUM NATURALIUM

U N I V E R S I TAT I S O U L U E N S I SACTAA

SCIENTIAE RERUM NATURALIUM

OULU 2017

A 687

Henni Ylänne

HERBIVORY CONTROL OVER TUNDRA CARBON STORAGE UNDER CLIMATE CHANGE

UNIVERSITY OF OULU GRADUATE SCHOOL;UNIVERSITY OF OULU,FACULTY OF SCIENCE;UNIVERSITY OF LAPLAND,ARCTIC CENTER

A 687

ACTA

Henni Ylänne

A C T A U N I V E R S I T A T I S O U L U E N S I SA S c i e n t i a e R e r u m N a t u r a l i u m 6 8 7

HENNI YLÄNNE

HERBIVORY CONTROL OVER TUNDRA CARBON STORAGE UNDER CLIMATE CHANGE

Academic dissertation to be presented with the assent ofthe Doctoral Training Committee of Technology andNatural Sciences of the University of Oulu for publicdefence in Kuusamonsali (YB210), Linnanmaa, on 24March 2017, at 12 noon

UNIVERSITY OF OULU, OULU 2017

Copyright © 2017Acta Univ. Oul. A 687, 2017

Supervised byDoctor Sari StarkProfessor Anne Tolvanen

Reviewed byProfessor Paul GroganProfessor Jukka Pumpanen

ISBN 978-952-62-1509-9 (Paperback)ISBN 978-952-62-1510-5 (PDF)

ISSN 0355-3191 (Printed)ISSN 1796-220X (Online)

Cover DesignRaimo Ahonen

JUVENES PRINTTAMPERE 2017

OpponentProfessor Richard Bardgett

Ylänne, Henni, Herbivory control over tundra carbon storage under climatechange. University of Oulu Graduate School; University of Oulu, Faculty of Science; University ofLapland, Arctic CenterActa Univ. Oul. A 687, 2017University of Oulu, P.O. Box 8000, FI-90014 University of Oulu, Finland

Abstract

Air temperatures in high-latitude regions are anticipated to rise by several degrees by the end ofthe century and result in substantial northward shifts of species. These changes will likely affectthe source and sink dynamics of greenhouse gases and possibly lead to a net carbon release fromhigh-latitude soils to the atmosphere. However, regional differences in carbon cycling dependhighly on the vegetation community composition, which may be controlled by the abundance ofherbivores. I investigated whether mammalian herbivores, mainly reindeer and rodents, alterecosystem carbon storage through their impacts on vegetation and on dominant plant functionaltraits. I combined observations of recent changes in ecosystem carbon with experimental fieldmanipulations of both herbivory and climate change and measured carbon storage in vegetationand soil, the uptake and release of carbon dioxide, microbial activity and compared these to plantcommunity composition.

Results of my PhD thesis show that under ambient conditions, the impacts of herbivory on bothabove- and belowground carbon storage ranged from positive to negative. Herbivory altereddominant plant functional traits and these were fairly good predictors of the changes in soil carbon.When combined with experimental warming, herbivory continued to exert control on thedominant plant functional traits but the strong effects of warming on ecosystem carbon storagemostly concealed the impact of herbivory. Interestingly, herbivory–nutrient interactions that werenot linked to dominant functional traits determined the consequences of warming on soil carbon.Taken together, I show clear and site-specific impacts of herbivores on vegetation and ecosystemcarbon storage and the processes that govern them. Therefore, I suggest that an improvedunderstanding of the role of herbivory in the global carbon cycle could improve estimations ofglobal carbon–climate feedbacks.

Keywords: carbon sequestration, CO2 flux, global warming, grazing, land-use, plantfunctional traits, Rangifer tarandus, reindeer, soil carbon

Ylänne, Henni, Herbivorian vaikutus tundran hiilivarastojen kehitykseenlämpenevässä ilmastossa. Oulun yliopiston tutkijakoulu; Oulun yliopisto, Luonnontieteellinen tiedekunta; Lapin Yliopisto,Arktinen keskusActa Univ. Oul. A 687, 2017Oulun yliopisto, PL 8000, 90014 Oulun yliopisto

Tiivistelmä

Vuosisadan loppuun mennessä arktisten alueiden lämpötilan odotetaan nousevan usealla asteel-la ja johtavan lajien siirtymiseen yhä pohjoisemmaksi. Nämä muutokset todennäköisesti muutta-vat pohjoisten ekosysteemien kykyä vapauttaa ja sitoa ilmakehän hiiltä ja saattavat johtaa sii-hen, että yhä enemmän hiiltä vapautuu tundramailta ilmakehään. Kuitenkin paikallisesti hiilen-kierto on riippuvainen kasviyhteisöstä ja erityisesti kasvien funktionaalisista ominaisuuksista.Väitöskirjassani tutkin, voivatko herbivorit, pääasiassa porot sekä jyrsijät, muokata hiilenkiertoamuuttamalla kasvillisuutta. Tutkimuksissani seurasin kuinka alueen laidunnushistoria on muo-kannut hiilivarastoja ja hiilenkiertoa tällä hetkellä ja pyrin arvioimaan herbivorien vaikutustalämpenevässä ilmastossa kokeiden avulla, joissa manipuloidaan sekä herbivoriaa että lämpötilaatai ravinteiden saatavuutta. Tulokseni perustuvat arvioihin hiilen varastoista, hiilidioksidinvapautumisesta ja sitoutumisesta sekä mikrobien aktiivisuudesta, joita vertaan kasviyhteisöön.

Tulokseni osoittavat, että herbivoria voi joko lisätä tai vähentää ekosysteemin hiilivarastojasekä maan päällä että maan alla. Muutokset hiilivarastoissa selittyivät varsin hyvin herbivorientuottamilla kasvillisuusmuutoksilla ja valtalajien funktionaalisilla ominaisuuksilla. Herbivoriamuokkasi kasviyhteisöä myös kokeellisen lämmityksen yhteydessä, mutta lämmityksen välittö-mät vaikutukset hiilivarastoihin peittivät suureksi osaksi alleen herbivorian vaikutukset. Kuiten-kin herbivorian ja lannoituksen kasvillisuusmuutoksista riippumattomat yhdysvaikutukset mää-rittivät lämpenemisen seuraukset maan hiileen. Kaiken kaikkiaan, tutkimukseni osoittaa, ettäherbivorit voivat paikkakohtaisesti muokata kasvillisuutta, ekosysteemin hiilivarastoja sekä hii-lenkierron prosesseja. Näiden tulosten myötä ehdotan, että parempi ymmärrys herbivorian vai-kutuksista maailmanlaajuisesti voisi parantaa nykyisiä ennusteita siitä, kuinka ilmaston lämpe-neminen muuttaa hiilenkiertoa.

Asiasanat: hiilen varastoituminen, hiilidioksidivuo, laidunnus, maan hiili, maankäyttö,poro, Rangifer tarandus

7

Acknowledgements

I wish to thank Sari Stark for taking me to the field as a graduate student and later

asking me along on this thesis project. I have felt privileged to be able to do this as

work – and I’ve been lucky for having you as a supervisor! These have been good

years, we have had good trips to the field and I’m especially grateful to you for

introducing me to great bunch of collaborators. I’m also thankful to Anne Tolvanen

for supervising me and being such an inspiring person to work with. Although we

have not been much in contact in the later years, each time we have met, your

positive feedback and the good working spirit have motivated me.

This project would not have been the same without the great co-authors and all

the help I have received in the field and in the laboratory. Johan, Maria, Elina, Lauri,

Minna and Saija: We have been doing cool research together and I’ve been inspired

by each of you. Johan, I always felt you had time for my questions. Maria and Elina,

your co-support has cheered many days during the last years! Lauri, I admire the

fairness of your working style, not least giving me and my crew place to stay

overnight in Alta and for letting me teach in your course.

During these four years both the Arctic Centre at the University of Lapland and

the Department of Ecology and Genetics at the University of Oulu have been really

helpful to me and great places to do research. Thanks especially to Bruce and

Annamari for smoothening the way at both places. I also wish to thank my great

co-workers and office mates at Oulu: I’ve met incredible people and have had a

great atmosphere of co-support around me all the time. Thank you for cheering up

so many days of mine!

I also want to thank professors Paul Grogan and Jukka Pumpanen for pre-

examining this thesis and inspiring me to think further. You might notice that this

is not the version that I sent you a few months ago. Pauls suggestion on improved

ways of presenting and analyzing the data of the second paper turned to paper to

something else and changed my interpretation of the results. Although, it was quite

a tough core in the given time, I’m having a good feeling of the outcome and I truly

appreciate the honesty and the input!

I wrote the introduction to this book with the aim to tell my friends and family

what I have been doing in the past years. Now I see that I got carried away but I

still hope the defence and parts of this book would shed light on my second life in

the tundra and in the office. I know I’ve been lucky with the good people around

me, motivated by the common plans on saving the planet and refreshed by the

shared excursions to the wilderness!

8

This work has been funded by Maj and Tor Nesslings Foundation and Kone

Foundation and has been conducted in the premises of the Department of Ecology

and Genetics in the University of Oulu, the Arctic Centre in the University of

Lapland and the Natural Resources Institute Finland in Rovaniemi. The Kilpisjärvi

Biological Station and Reisadalen Hytteutleje in Sappen have are also highly

appreciated for the great accommodation during the field campaigns.

Oulussa 10.2.2017 Henni Ylänne

9

Original publications

This thesis is based on the following publications, which are referred throughout

the text by their Roman numerals:

I Ylänne H, Olofsson J, Oksanen L & Stark S (Manuscript) Grazer-induced grassification of deciduous shrub tundra may promote soil carbon sequestration.

II Ylänne H, Kaarlejärvi E, Väisänen M, Männistö M, Ahonen S, Olofsson J & Stark S (Manuscript) Fertilization mitigates warming-induced carbon loss in long-term grazing regimes but not in short-term exclosures.

III Ylänne H, Stark S, & Tolvanen A (2015) Vegetation shift from deciduous to evergreen dwarf shrubs in response to selective herbivory offsets carbon losses: evidence from 19 years of warming and simulated herbivory in the sub-arctic tundra. Glob Chang Biol 21: 3696–3711.

10

11

Contents

Abstract

Tiivistelmä

Acknowledgements 7 Original publications 9 Contents 11 1 Introduction 13

1.1 The influence of vegetation community composition on carbon

cycling ..................................................................................................... 14 1.2 The control of tundra herbivores over ecosystem processes ................... 16 1.3 Previous evidence of herbivore effects on tundra carbon storage ........... 18 1.4 Aims of the study .................................................................................... 19

2 Materials and methods 21 2.1 Study areas .............................................................................................. 21 2.2 Experimental designs .............................................................................. 23 2.3 Field sampling and laboratory procedures .............................................. 25 2.4 Statistical analyses .................................................................................. 28

3 Results and discussion 29 3.1 Evidence from herbivores altering ecosystem carbon stocks

under current climate .............................................................................. 29 3.2 Interactive effects of herbivores and climate change on

ecosystem carbon .................................................................................... 35 4 Conclusions 41 List of references 43 Original publications 51

12

13

1 Introduction

For millennia, cold and waterlogged conditions in the arctic and subarctic areas

have limited microbial activity more than the uptake of carbon by vegetation. This

has resulted in the storage of up to 67 Pg of carbon in arctic and subarctic soils

(Tarnocai et al. 2009) – more than twice the amount of carbon currently in the

atmosphere. Most of this carbon is in permafrost soils and arctic wetlands but also

tundra soils with only sporadic permafrost may store up to 100 g C m-2 (Schuur et al. 2015). Now that average temperatures are rising twice as fast in the Arctic than

elsewhere in the world, north-ward shifts of species or even whole biomes are

anticipated (IPCC 2014). These changes, both in temperature and vegetation, are

predicted with high confidence to alter the rates of carbon uptake and release by

ecosystems, the phenomenon referred to as carbon–climate feedback. The

magnitude of this feedback is still uncertain (Schuur et al. 2015, Crowther et al. 2016), despite its crucial role for determining the limits for greenhouse gas

emissions to meet climate targets.

The general prediction is that climate change will likely exert a positive effect

on plant productivity and increase plant carbon reserves (Lu et al. 2013). However,

also the decomposition of soil carbon is likely to accelerate as soil microbial

activity increases with warmer temperatures and drier soils (Davidson & Janssens

2006), with enhanced litter quality (Cornwell et al. 2008) and with altered

microbial communities (Karhu et al. 2014). For the whole permafrost region,

carbon losses are anticipated to outpace carbon gains by the year 2100 (Schuur et al. 2015, Crowther et al. 2016). In Fennoscandia and other areas with only sporadic

permafrost, the balance between carbon release and sink capacity may be more

subtle. Moreover, increasing evidence suggests that the regional patterns of carbon

uptake and release are controlled by the dominant vegetation types (Nobrega &

Grogan 2008, Cahoon et al. 2012a, Marushchak et al. 2013, Parker et al. 2015).

In this thesis I investigate whether mammalian herbivores, mainly reindeer and

rodents, can control ecosystem capacity to store carbon through their influence on

vegetation. Firstly, I will discuss how a vegetation community may affect

ecosystem carbon storage and introduce the concept of plant functional traits.

Secondly, I will give an overview on the herbivores of the tundra and present

pathways of how herbivores affect vegetation dynamics and ecosystem processes

– and ultimately, the tundra carbon cycle.

14

1.1 The influence of vegetation community composition on carbon

cycling

Individual plant species differ in how they influence ecosystem processes. To

incorporate the influence of each plant species on ecosystem functioning is,

however, challenged by the fact that there are more than 350 000 plant species

globally and about 2 000 in the tundra (The Plant List 2013). To simplify this, plant

species have been classified based on their functional traits, e.g. the common

attributes of species in responding to environmental changes and in affecting the

environment (e.g. Chapin et al. 1996). In its simplest form, the classification

separates traits based on plant growth forms, e.g. deciduous shrubs, evergreen

shrubs, forbs, grasses, sedges, bryophytes and lichens (Chapin et al. 1996), which

I present in Figure 1. This division is a coarse simplification and way more

functional traits may be used to describe differences within growth forms in litter

quality (Cornwell et al. 2008) and ultimately, soil carbon sequestration (Deyn et al. 2008). These functional traits would also include classification of mycorrhizal

status (Averill et al. 2014) and other root-driven processes (Bardgett et al. 2014).

Still, even the simple division of growth forms is a relatively accurate predictor of

litter quality and the associated litter feedbacks on tundra ecosystems (Dorrepaal

2007).

The cryptogams, e.g. bryophytes and lichens, are considered to increase soil

carbon sequestration because they produce litter of extremely poor quality, insulate

soils, trap water and nutrients and release compounds that are toxic to microbes and

other plants (Wal & Brooker 2004, Cornelissen et al. 2007, Gornall et al. 2007,

Stark et al. 2010). On the contrary, vascular plants have increased access to soil

water, nutrients and light through their root and stem systems. With faster growth

rates and higher nutrient concentrations they generally produce less recalcitrant

litter and have higher metabolic activity resulting in losses of carbon through

microbial and plant activity (Chapin et al. 1996, Deyn et al. 2008). Still, vascular

plants may also have recalcitrant tissues, such as the lignin-rich and nitrogen-poor

stems of deciduous and evergreen shrubs, evergreen leaves with high

concentrations of secondary defence compounds or the root litter in general (Deyn

et al. 2008, Rasse et al. 2005, Freschet et al. 2013).

15

Fig. 1. Hierarchical classification of plant growth forms in tundra (sensu Chapin et al. 1996) and their effects on soil carbon sequestration (Chapin et al. 1996, Deyn et al. 2008,

Freschet et al. 2013).

16

Chapin et al. (1996) include a level to the division of plant functional traits that

divides vegetation to low-stature tundra vegetation and tall woody vegetation.

Through protruding above the snow the taller woody vegetation decreases winter-

time albedo, increases sensible heat flux, is more prone to fires, has higher annual

carbon intake and demands more water than low stature vegetation (Myers-Smith

et al. 2011). Therefore increased range and biomass of woody shrubs within tundra

could have drastic effects on ecosystem functioning and carbon storage with

feedbacks on climate (Myers-Smith et al. 2011, Hartley et al. 2012, Bonfils et al. 2012). This raises concern over recent observations of expanded ranges and higher

biomass of tall-stature shrubs across the circumpolar tundra, seen also within

several experiments simulating climate change (Myers-Smith et al. 2011, 2015;

Elmendorf et al. 2012a, b).

1.2 The control of tundra herbivores over ecosystem processes

Although primary production is relatively low in the tundra, a multitude of

vertebrate herbivores thrive on Arctic areas (Barrio et al. 2016). The coastal areas

are mostly under the influence of avian herbivores, such as geese and swans,

whereas further inland, there are more species of mammalian herbivores, such as

hares, lemmings and voles (Barrio et al. 2016). There are six species of large

mammalian herbivores in the Arctic. These are mostly bovids and deer, with the

exception of wood bison in North America (Nowak 1999). The most widely

distributed of these is the reindeer and its subspecies caribou (Rangifer tarandus)

that graze much of northernmost Eurasia and North America (Bernes et al. 2015).

In Eurasia, most reindeer herds are semi-domesticated and their numbers and

grazing patterns have been strongly regulated by humans for thousands of years

(Forbes & Kumpula 2009). Besides the wild and semi-domesticated herbivores also

cattle, horses and sheep have been kept for centuries in arctic areas of high human

influence (Dallimer et al. 2009).

Herbivores have three direct means of impacting ecosystems: removal of plant

biomass, trampling of vegetation and soil and excretion of faeces and urine with

high nutrient concentrations (Fig. 2). These alter plant communities because

herbivores often select for species with high nutritious value and because plant

species differ in their ability to tolerate both grazing and trampling and to use the

increased available nutrients (Bardgett & Wardle 2003). As a result, herbivory may

provide a competitive advantage for plant traits that tolerate disturbances, grow fast

and produce nutrient-rich litter (Wardle et al. 2004). Alternatively, under conditions

17

Fig. 2. Different pathways of herbivore control over ecosystems. Herbivores directly

impact vegetation and soil through grazing, trampling and faecal fertilization. Herbivory

impacts on vegetation vary with plant species (in this figure graminoids and shrubs) as

grazing may be selective and plant species differ in their tolerance to trampling. This

affects the competition between plant species and the quality and quantity of the

produced litter. The vegetation changes feedback to soils, due to differences in biomass

allocation to roots, the production of root exudates and changes in the soil physical

environment (e. g. the influence of shading or water-trapping). The belowground

processes, once again, feedback to the aboveground environment because plant

species respond differently to the altered soil environment. In this figure the dotted and

even lines represent the varying impacts on / of different plant species.

with low nutrient availability, herbivory might favour plant species with high

concentrations of secondary defence compounds that are avoided by herbivores and

retard nutrient cycling rates (Wardle et al. 2004).

Herbivores may alter the abiotic environment of plants: trampling generally

increases the compaction of soil layers and the faeces increase the availability of

inorganic nutrients (Stark et al. 2008) which may accelerate nutrient cycling rates

(Olofsson et al. 2001, 2004, Stark et al. 2002). The abiotic changes may also be

induced by plant community changes: herbivory-induced decrease in the cover and

thickness of bryophyte or lichen layers decreases the insulation of soil (Wal &

Brooker 2004, Gornall et al. 2009, Susiluoto et al. 2008) and an herbivory-induced

decrease in vegetation and litter cover decreases the shading of soil (Elliott & Henry

2011). Herbivory may alter the resource inputs from vegetation to the soil by

altering the physiology of vegetation or through changes in vegetation community

18

(Bardgett & Wardle 2003). Through these changes in abiotic conditions and in

plant–plant, plant–microbe and microbe–microbe interactions, herbivores also

likely alter processes within the ecosystem, such as the cycling of nutrients and

carbon (Wardle et al. 2004, Stark & Väisänen 2014) – however, in controversial

directions (Bardgett & Wardle 2003).

Ultimately, continued high abundance of herbivores may lead to a shift into an

alternative ecosystem state (Beisner et al. 2003). The most-known evidence from

herbivore-induced ecosystem state transition comes from the savanna, where large

mammals limit tree recruitment and keep vast savanna areas open (Scholes &

Archer 1997). Also in the tundra, herbivory could lead to stepwise vegetation

transitions in which the productivity and hence, the herbivore-carrying capacity,

increases with grazing pressure (Wal 2006). Such herbivory-induced alternative

vegetation states are mostly found along natural borders of grazing (Olofsson et al. 2001, 2004) and supported by paleo-ecological studies, where megafauna are

suggested to have driven large-scale ecosystem transitions in the late Pleistocene

(Zimov et al. 1995, Barnosky et al. 2016).

1.3 Previous evidence of herbivore effects on tundra carbon

storage

Even though several studies have suggested the role of herbivores in controlling

ecosystem carbon storage (Tanentzap & Coomes 2012, Schmitz et al. 2014),

relatively few studies have been conducted to test this. These show that the impact

of herbivores on ecosystem carbon are by no means consistent and herbivores may

increase (Martinsen et al. 2011, Hafner et al. 2012), decrease (Cahoon et al. 2012b¸

Metcalfe & Olofsson 2015) or lead to no changes in ecosystem carbon (Susiluoto

et al. 2008, Köster et al. 2015).

In Greenland and Fennoscandia, the herbivory exclosures have shown to

increase net ecosystem CO2 uptake (Cahoon et al. 2012b, Metcalfe & Olofsson

2015) and also decadal history of light grazing had led to higher CO2 uptake

compared to adjacent heavily grazed areas (Väisänen et al. 2014). The analogous

outcomes in these studies were, however, driven by different mechanisms: in the

studies using experimental herbivore exclosures, herbivores were found to suppress

plant productivity and shrub growth (Cahoon et al. 2012b, Metcalfe & Olofsson

2015), whereas when comparing sites with a contrasting decadal history of

herbivory, the higher herbivory intensity had increased ecosystem respiration to a

greater extent than productivity (Väisänen et al. 2014). Although these results and

19

several others (Speed et al. 2010, Falk et al. 2015) indicate a negative impact of

herbivores on ecosystem carbon storage, moderate herbivory intensities have been

shown to increase soil carbon stocks when compared to herbivory exclosures

(Martinsen et al. 2011, Hafner et al. 2012, Sjögersten et al. 2012).

1.4 Aims of the study

The aim of this study was to examine how climate change and herbivores alter

vegetation structure and the capacity of subarctic tundra ecosystems to store carbon.

More specifically I will ask the following the questions:

1. What are the recent trends (14 years) in tundra ecosystem carbon stocks and

do the trends depend on the intensity of herbivory?

(Paper I)

2. What are the expected, climate change-induced, changes in tundra carbon

stocks under contrasting intensities of herbivory?

(Papers II, III)

3. Do the changes caused by herbivory and climate in ecosystem carbon stocks

correlate with the changes in the plant functional types, soil abiotic properties

and microbial community composition?

(Papers I, II, III)

4. How are the changes in carbon stocks reflected in ecosystem processes, such

as the release and fixing of CO2, microbial respiration and the quantity and

quality of microbial extracellular enzymes?

(Papers I, II, III)

To answer these questions, I quantify the impact of herbivores through comparison

of different herbivore intensities (I, II), by using short-term exclosures (II) and by

simulating herbivore damage through clipping (III). In the first paper, I compared

contrasting herbivore intensities along two reindeer pasture rotation fences. In the

second paper, I compared the responses to experimental warming and fertilization

along one of the fences, where also short-term reindeer exclosures were built. In

the third paper, I compared how simulated herbivory on the dominant plant species

interacts with the responses of experimental warming.

20

I formed predictions on the effects of herbivory and climate change based on

the theory that different plant functional traits control soil carbon sequestration

(sensu Deyn et al. 2008). I hypothesized that by altering the relative abundance of

plant growth forms, herbivores could alter the rates of carbon sequestration and –

ultimately – the amount of carbon stored in ecosystems (Schmitz et al. 2014).

Generally, in the absence of herbivores (I, II, III), I expected to see an increase in

the abundance of deciduous shrubs in response to climate warming (Myers-Smith

et al. 2015). This was predicted to result in higher aboveground carbon stocks, but

the faster cycling of nutrients and carbon in soil (Nobrega & Grogan 2007, Deyn

et al. 2008, Parker et al. 2015) was predicted to decrease soil carbon stocks (Mack

et al. 2004). Under high herbivory intensity (I, II), where herbivory numbers had

increased over time (Fig. 2 in I), I expected to see increased dominance of

graminoids in the vegetation (Olofsson et al. 2001, 2004). As herbivory may

increase the allocation of graminoids’ resources to root biomass (Gao et al. 2007,

Hafner et al. 2012) and as graminoids produce relatively recalcitrant root litter

(Rasse et al. 2005, Freschet et al. 2013), I predicted that the increased dominance

of graminoids would increase soil carbon sequestration. Under the combined

warming and herbivory treatment (III) and under the warming treatment within

short-term exclosures (II), I expected evergreen shrubs to increase in abundance,

as reported earlier on the sites (Rinnan et al. 2009, Väisänen et al. 2014). Because

evergreen shrubs grow slowly and produce more recalcitrant litter compared to

deciduous shrubs (Deyn et al. 2008), this was predicted to result in lower rates of

gross ecosystem production and ecosystem respiration and result in weaker effects

of warming on ecosystem carbon.

21

2 Materials and methods

In this thesis, I use the evidence from field monitoring and experimental

manipulations to detect the interactive effects of climate change and herbivory on

subarctic carbon stocks. Firstly, I monitored how vegetation and ecosystem carbon

stocks have changed over the past 14 years alongside warmer temperatures. The

study covered three zones that had experienced contrasting herbivore intensities for

the past 50 years. This enabled me to compare how different decadal histories have

affected current carbon stocks and processes and to analyse whether recent trends

differed according to herbivory intensity. Secondly, I used two existing field

experiments that combine herbivory with warming or with warming and enhanced

nutrient availability. Although, I could not compare the development of carbon

stocks over the experimental time, the differences between treatments after 19 and

5 years of the experiments provide information on how herbivory interacts with

ameliorated warming in shaping vegetation and ecosystem carbon storage.

2.1 Study areas

The studies were commenced in northern Norway and Finland on three tundra

heaths: Raisduoddar (69º31’29 N, 21º19’16 E), Čearro (69°43’23 N, 21°37’45 E)

and Kilpisjärvi (69º04’05 N; 20º49’11 E, Fig. 3). Raisduoddar and Kilpisjärvi are

100–200 m above the treeline on the slope of the mountains Gahpperus and Jehkats,

whereas Čearro is located on a highland plateau. Among the sites, Raisduoddar

represents the most fertile ecosystem with a mixture of evergreen and deciduous

shrubs of Betula nana, Empetrum nigrum spp. hermaphroditum and Salix species.

The site in Kilpisjärvi is on a snowbed with Vaccinium myrtillus as the dominant

species. In Čearro the vegetation is composed mainly of B. nana and the ground

layer is rich in Cladonia-lichens.

All sites are grazing grounds of the semi-domesticated reindeer (Rangifer tarandus tarandus L). The sites in northern Norway, Raisduoddar and Čearro are

located along reindeer pasture rotation fences that separate the coastal summer

ranges of reindeer from the winter ranges inland. Reindeer arrive at both sites in

early August and stay in the vicinity of the fence on the summer range for two to

three weeks. The other side of the fence on the autumn-spring range is used mainly

as a by-pass corridor, resulting in a clear and visible difference in herbivore

intensity. Decades of heavy grazing by reindeer on the summer range have shifted

the vegetation from shrub dominance towards a dominance of graminoids

22

(Oksanen et al. 1978, Olofsson et al. 2001, 2004), accelerated the nutrient cycling

rate (Stark et al. 2002, Stark & Väisänen 2014) and increased the dominance of

bacteria in the soil microbial community (Männistö et al. 2016). Because reindeer

tend to follow boundaries, areas further away from the fence are subjected to a

moderately lower grazing intensity (Olofsson et al. 2001, 2004).

The Kilpisjärvi site is situated on the year-round reindeer grazing regime of the

Käsivarsi reindeer herding district. In contrast to the other sites, the effects of

herbivores were tested experimentally by clipping, which simulates the activity of

particularly voles and lemmings typical of the snow-protected depressions and

slopes (Virtanen et al. 1997, 2002).

Fig. 3. Reindeer grazing ranges in northern Norway and the location of the study sites.

The coastal reindeer summer ranges are marked in dark grey, whereas the autumn,

winter and spring ranges are further inland. The study site in Finland (Kilpisjärvi) is on

a year-round grazing range.

23

2.2 Experimental designs

Resampling study under contrasting herbivory intensities (I)

I investigated recent temporal changes in ecosystems under contrasting herbivory

intensities by analysing ecosystem carbon stocks along two 50-year-old reindeer

pasture rotation fences in Raisduoddar and Čearro in 2014 and compared these with

previous data collected in 2000 (I, Olofsson et al. 2004). The experimental setup

consisted of eight transects spaced 100 m apart that extended through three grazing

intensities. Light grazing intensity (LG) was defined as the zone 10 m from the

fence on the autumn-spring range, heavy grazing (HG) as the zone 10 m from the

fence on the summer range and moderate grazing (MG) as the zone 100 m from the

fence also on the summer range. Olofsson et al. (2004) demonstrate a significant

difference in the activity of reindeer among the grazing intensities and the impact

of grazing intensity on vegetation is also visible (Fig. 4). The eight transects yielded

a total of 24 study plots at both sites.

It was not possible to determine the mechanisms that underlie the changes in

carbon stocks over the past 14 years as these may be driven concurrently by several

factors: by warmer temperatures (Fig. 2a in I), by higher reindeer numbers since

2000 (Fig. 2b in I) and by transient dynamics since the building of the pasture

rotation fences (as described in Olofsson et al. 2001, 2004). Therefore, this

temporal comparison of ecosystem carbon stocks is rather used to assess how the

observed changes in vegetation influence soil carbon sequestration under

contrasting herbivory intensities. Moreover, I expected this study to provide

insights into the relationship between ecosystem carbon and the dominant plant

functional traits that were altered by herbivores (Olofsson et al. 2001, 2004).

Experimental warming and fertilization under contrasting long- and short-

term herbivory intensities (II)

I tested the interactive effects of herbivory, experimental warming and increased

nutrient availability on carbon stocks in a full-factorial design that was

implemented along the 50-year-old reindeer fence in Raisduoddar in the year 2010

(Väisänen et al. 2014). The experiment consisted of the treatments: control,

warming, fertilization and combined warming and fertilization that were replicated

on the lightly grazed (LG) and the heavily grazed (HG) sides of the fence. We also

24

built short-term exclosures under heavy grazing (HGexc) to surround one half of

each study plot to test recovery from grazing.

Experimental warming was simulated with open-top chambers (OTCs, Marion

et al. 1997, Fig. 5) that increase surface temperatures on average by 1.5–1.9 ºC and

soil temperatures by 0.6–1.1 ºC (Elmendorf et al. 2012a, Lu et al. 2013). OTCs also

decrease wind speed and soil moisture and increase air humidity (Bokhorst et al. 2011). Yet, the vegetation changes within OTCs are closely analogous with reports

from natural warming (Hollister & Webber, 2000, Elmendorf et al. 2012a, b) which

gives support to using OTCs as a proxy for the consequences of climate warming.

Fertilization was implemented by adding 10 g N m-2 in the form of ammonium

nitrate (NH4NO3) to the study plots. This dosage corresponds to the predicted soil

nitrogen increase after a 7 ºC increase in air temperature (Mack et al. 2004).

In the experiment, the short-term exclosures were built only for the time of

reindeer migration. OTCs were built after snow-melt and removed prior to the

reindeer migration, which resulted in two months of warming in the beginning of

the growing season each year. Fertilization was applied in the beginning of the

growing season, on average in late June, when the soil was no longer frozen. The

four treatments under the three grazing intensities were repeated eight times along

the fence resulting in 96 study plots altogether. This study was sampled for carbon

stocks in 2014 after five summers of the experiment. The activities of microbial

extracellular enzyme activities, the abundance of microbes and the soil

fungal:bacterial ratio were analysed a year earlier, in 2013.

Experimental warming and simulated herbivory (III)

I tested the interactive effects of simulated herbivory and warming on ecosystem

carbon stocks and CO2 fluxes in a full-factorial experiment set to Kilpisjärvi in

1994 (Rinnan et al. 2009). Warming was simulated by open-top chambers that were

kept on the study plots year round. Simulated herbivory was implemented on the

dominant species, Vaccinium myrtillus, by clipping every second leaf during the

peak of the growing season each year. This manipulation allowed standardized

treatments between years, but has its disadvantages as it does not simulate the real

feeding patterns of herbivores and it excludes completely herbivory-impacts via

trampling and fertilization by urine and faeces. The experimental treatments were

repeated five times resulting in 20 study plots. This study was sampled in year 2013

after 19 years of treatment.

25

Fig. 4. The impact of herbivory on vegetation in Čearro. From left to right, the figures

represent typical vegetation under light, moderate and heavy grazing.

Fig. 5. Open-top chambers were used to simulate warmer growing seasonal

temperatures in Raisduoddar and Kilpisjärvi. The chamber on the right has received a

combined warming and fertilization treatment.

2.3 Field sampling and laboratory procedures

Carbon stocks and vegetation abundance

From the study plots, I collected biomass of vascular aboveground vegetation,

cryptogams, litter, roots as well as organic and mineral soil layers and determined

the concentration of carbon and nitrogen in these (I, II, III). I collected vegetative

aboveground biomass by clipping the aboveground parts of vascular plants and

pooled these later in plant growth forms (Table 1). I took composite cores of the

26

ground layer and handpicked bryophytes, lichens and litter from these. I estimated

organic and mineral soil biomass by coring 3–7 soil samples from each plot. These

were separated for the layer of organic and mineral soil, combined with each other

and sieved in the laboratory. I separated roots from the sieving residues and washed

these. All samples were dried and weighed before homogenizing them for the

carbon and nitrogen analysis. As the area of each sample was known, I counted the

carbon and nitrogen stocks per area. I conducted this sampling once in the middle

of the growing season in either 2013 (III) or in 2014 (I, II), before the herbivory

simulation (III) or reindeer arrival to the areas (I, II). In Raisduoddar and Čearro,

the same procedure had been used in 2000 (Olofsson et al. 2004, but see differences

in I: Methods), which enabled detection of recent changes in ecosystem carbon

stocks.

To see how carbon stocks and processes depend on vegetation community

composition, I monitored vegetation abundance at the peak of the growing season.

To ensure comparison with previous measurements, the method varied slightly

between the studies: in Studies I and II, I used point-intercept analyses with 10–40

pins (Jonasson 1988), in Study III, a point-frequency method on 100 systematically

placed points was used (Knapp 1984).

Ecosystem CO2 exchange

In the studies I and III, I measured gross ecosystem productivity and ecosystem

respiration, which together constitute the net ecosystem CO2 exchange. To depict

seasonal variation in the fluxes, the measurements were repeated on even intervals

throughout the growing season. I used a custom-built chamber (Tunkua Oy,

Finland) within which I recorded the change in CO2 concentration, temperature,

humidity and photosynthetic active radiation over 90 s. On each plot, I made several

consecutive measurements of the CO2 change by gradually decreasing the light

intensity with hoods made of white mosquito nets. During the last measurements,

the chamber was covered with an opaque hood that prevented light from reaching

the plots giving an estimation of ecosystem respiration.

I calculated gross ecosystem productivity as the difference between net

ecosystem exchange and ecosystem respiration. From the combination of several

light measurements with different shading levels, I derived a light response curve

of photosynthesis for each plot on each occasion (Williams et al. 2006). This light

response curve was used to standardize the CO2 flux to a common irradiance (600

µmol photons m-2 s -1). In Study I, I calculated the average mid-day CO2 flux rates

27

over the growing season as weighted means that take into account the uneven

intervals between the measurement occasions. In Study III, I used interpolated data

on hourly irradiance and air temperatures for the site (provided by Finnish and

Swedish Meteorological Institutes) and modelled the fluxes for each hour of the

growing season based on changes in leaf area, the light response of vegetation and

the temperature-dependency of ecosystem respiration (the model descriptions were

derived from Shaver et al. 2007). I tested the fit of the modelled values by making

additional flux measurements outside of midday in the beginning, the middle and

the end of the growing season.

Microbial respiration

In Studies I and III, microbial respiration was analysed in the laboratory on monthly

intervals during the growing season. Within a week of field sampling, subsamples

of sieved fresh soil were put into sealed incubation bottles into standardized

temperature. After three days, the bottles were once aerated to enable

acclimatization of soils after the sieving disturbance. The bottles were again placed

in the standardized temperatures for 72 hours, after which air samples were

collected from the headspace of the bottles and analysed for CO2 concentrations

with a gas chromatograph. In Study I, the procedure was repeated at four

temperatures (4, 9, 14 and 19 °C) to depict the temperature response of microbial

respiration (Wallenstein et al. 2009).

Soil parameters, microbial activity and community composition

I examined soil temperature variation throughout the growing season by installing

soil temperature loggers at the study plots (I, II, III). I also analysed soil pH,

moisture, organic matter content and the concentration of inorganic nutrients from

the same soil samples that were used for soil carbon content or microbial activity

analyses. In Study II, I incorporated the data to analyses of soil extracellular

enzyme activities and microbial community composition. Microbial activity was

quantified as the potential activity of β-glucosidase, phenol oxidase and N-acetyl-

glucosaminidase. β-glucosidase depolymerizes cellulose and is used as a proxy for

soil carbon turnover (Allison et al. 2010). Phenol oxidase catalyses the breakdown

of lignin and other aromatic compounds and is linked to the decomposition of

recalcitrant carbon compounds (Sinsabaugh 2010). N-acetyl-glucosaminidase

hydrolyses fungal chitin and is linked to microbial turnover rates (Beier &

28

Bertilsson 2013). By labelling these with enzyme-specific substrates, the amount

of each enzyme in soil was quantified. This method does not indicate whether the

enzymes have been actively catalysing reactions in soil – therefore, their amount in

soil is referred to as the potential activity of soil microbial communities.

Microbial community composition was estimated by quantitative PCR, where

the amount of certain fungi- or bacteria-specific DNA-sequences (ITS2 and 16S,

respectively) was counted from frozen samples of sieved organic soil (Männistö et al. 2016). As both fungal and bacterial species differ in the number of sequences

within individuals, this method does not provide information on the total abundance

of fungi or bacteria in soils. Rather, it tentatively indicates shifts in community

composition towards certain species with either a high or low amount of DNA

sequences (Ihrmark et al. 2012). I also used this data to depict changes in the

relative abundance of fungi compared to bacteria (Fungal:bacterial ratio).

2.4 Statistical analyses

For statistical analysis, I mainly used the analysis of variance (I, III) with a repeated

measures design when needed. In Paper I, I defined transect as a random factor.

The tests were followed by Duncan’s or Tukey’s post hoc tests to spot the pair-wise

differences. I also monitored how the measured variables correlated with vegetation

community composition through non-metric multidimensional scaling (I, III). In

Paper II, I wanted to compare separately the impact of long-term herbivory (LG vs

HG) and short-term exclosures (HGexc vs HG). Therefore, I used a mixed model

for multilevel designs, where contrasts were modified so that both LG and HGexc

are compared to only HG. As random factor I included the block, the block within

either the summer or winter grazing range, and the pair of plots that were separated

by the short-term exclosures. All these variables were nested within each other. In

case of significant interactions, I ran post-hoc tests to spot the pair-wise differences.

All statistical tests were done with R statistical software (R development core

team, 2013) using the package nlme for linear mixed effect models (Pinheiro et al. 2014) and the package vegan for ordination (Oksanen et al. 2015).

29

3 Results and discussion

3.1 Evidence from herbivores altering ecosystem carbon stocks

under current climate

Herbivory-dependent changes aboveground

My results show increased vegetative biomass in Raisduoddar and Čearro over the

past 14 years (I) in line with general trends of increased vegetative biomass across

tundra (Elmendorf et al. 2012b). Also in both field experiments, Raisduoddar and

Kilpisjärvi, the abundance of vascular vegetation increased in control plots over the

past four and 19 years (II and III). Nevertheless, herbivory exerted profound effects

on the vegetation community at all sites, but to various directions.

In Raisduoddar, the mosaic of evergreen and deciduous shrubs under light

grazing was replaced by graminoids under heavy grazing, as was already reported

in 1997 and 2000 (Olofsson et al. 2001, 2004, I). In 2014, this grassification had

extended further away from the pasture rotation fence to the moderately grazed

zone. The short-term exclosures under heavy grazing increased vegetation cover,

particularly graminoids and decreased the amount of bryophytes (II). The lightly

and moderately grazed area in Čearro and the control plots in Kilpisjärvi were

dominated by deciduous shrubs, but herbivory shifted the vegetation to different

directions depending on site: in Čearro heavy grazing increased graminoids

(Olofsson et al. 2004, I), whereas in Kilpisjärvi, evergreen shrubs increased with

simulated herbivory (Rinnan et al. 2009, III). These vegetation changes were

concurrent with my hypothesis and are summarized in Figure 6.

In Raisduoddar and Čearro, herbivory exerted high control over the cover and

biomass of bryophytes. In Raisduoddar, both heavy and moderate grazing reduced

bryophyte biomass (I) and bryophytes did not recover in the short-term exclusion

(II). In Čearro, the lichen-rich ground layer that prevailed under light grazing was

replaced by bryophytes under moderate grazing, whereas under heavy grazing the

bryophytes to most extent disappeared. These impacts on the ground layer are

closely in line with the theory of alternative vegetation states by Wal (2006), where

lichen-rich grounds become dominated by bryophytes and ultimately by

graminoids when herbivory intensity is increased. In Čearro, all three vegetation

states were found and in Raisduoddar, only the latter two. Similar sensitivity of

lichen to reindeer grazing is reported across the tundra (Bernes et al. 2015), but

30

only sometimes these become replaced by bryophytes (Väre et al. 1995, Köster et al. 2015). The simulated herbivory treatment in Kilpisjärvi did not alter bryophyte

cover or biomass (Rinnan et al. 2009, III).

Concurrently with the changes in vegetation community composition,

herbivory altered the aboveground carbon stocks – but to contrasting directions

depending on the study. In Raisduoddar and Čearro, decades of high herbivore

intensity decreased the aboveground carbon stocks (I, II) in line with the

generalization that herbivores decrease aboveground carbon through biomass

consumption (Tanentzap & Coomes 2012). This is in line with predictions based

on the dominant plant functional traits as graminoids store considerably less carbon

than woody evergreen or deciduous shrubs (Deyn et al. 2008). Contrasting the

results from Raisduoddar and Čearro, the increase in evergreen shrubs under

simulated herbivory in Kilpisjärvi increased vegetative carbon stocks (III). This

demonstrates that herbivory may not always decrease vegetative carbon stocks

(Tanentzap & Coomes 2012). Rather, the direction of the effect of herbivory

depends on whether the shift in vegetation is towards plant species that store more

or less carbon in their aboveground biomass.

Contrasting patterns in soil and ecosystem carbon storage depend on

plant functional traits

In line with my hypothesis, the herbivory control over the dominant plant functional

traits was a rather good predictor of the effect of herbivory on soil and ecosystem

carbon storage. At the three sites, the effects of herbivory on soil carbon stocks

varied from negative to neutral and positive. In Kilpisjärvi, simulated herbivory

decreased soil carbon sequestration, but interestingly, the reduced soil carbon

stocks were compensated by increased vegetative carbon stocks. In Raisduoddar,

neither the short-term exclosures nor the long-term difference in herbivory led to

changes in ecosystem carbon stocks whereas in Čearro, herbivory significantly

increased soil sequestration and ecosystem carbon.

In Raisduoddar and Čearro, the heavily grazed, graminoid-dominated areas

converged in their ability to accumulate carbon (I). The ecosystem carbon stocks

differed between the two sites under light grazing, which could partly explain why

herbivory led to no changes in Raisduoddar and increased ecosystem carbon

storage in Čearro. I suggest that the difference in the carbon stocks between the

lightly grazed areas in Raisduoddar and Čearro results from the plant species traits

that dominate under light grazing: the mixture of E. hermaphroditum and B. nana

31

in Raisduoddar represents a shrub-heath ecosystem that stores more soil carbon

than deciduous shrub-tundras dominated solely by B. nana (Parker et al. 2015),

such as Čearro. These two shrubs have different mycorrhizal associates: E. hermaphroditum has ericoid mycorrhizal fungi and B. nana ectomycorrhizal

symbiotic fungi (Clemmensen et al. 2013). These have been suggested to retard or

stimulate, respectively, soil carbon cycling both in boreal forests (Clemmensen et al. 2013, Averill et al. 2014) and in tundra (Parker et al. 2015). My results reveal

that, in their ability to sequester carbon, the graminoid-dominated tundra could fall

between the evergreen-ericoid mycorrhizal communities and the deciduous-

ectomycorrhizal communities (Fig. 7). I suggest that the reason behind the

relatively high carbon sequestration in the graminoid-dominated tundra stems from

the graminoids’ high investment of resources in root biomass, especially when

grazed (Gao et al. 2007, Hafner et al. 2012). This increases soil carbon storage

because graminoid root litter is relatively recalcitrant (Rasse et al. 2005, Deyn et al. 2008, Freschet et al. 2013).

My studies show two cases where herbivory drove the depletion of deciduous

shrubs (Čearro and Kilpisjärvi). These cases differed in the vegetation that replaced

the deciduous shrub-dominated vegetation and had contrasting consequences on

ecosystem carbon: In Čearro, where the deciduous B. nana was replaced almost

totally by graminoids, soil carbon stocks increased (I, Fig. 7). In Kilpisjärvi, where

evergreen shrubs increased with simulated herbivory, soil carbon stocks decreased

(III, Fig. 7). It is noteworthy that in Kilpisjärvi, the decrease in soil carbon stocks

occurred concurrently with an increase in aboveground carbon stocks and led to no

changes in ecosystem carbon. I suggest that the different outcome results from

limitations of the simulated herbivory treatment. In Čearro, heavy grazing intensity

increased soil nitrogen stocks and soil NH4-N concentrations (Olofsson et al. 2004,

I), concurrent with the theory that herbivores increase both nutrient availability and

decomposition rates (e. g. Stark et al. 2002, Stark & Väisänen 2014). Nevertheless,

in Kilpisjärvi the simulated herbivory tentatively decreased soil nitrogen stocks (P

= 0.07). These results suggest that the fertilizing effects of herbivore urine and dung

on the soil are crucial for determining ecosystem responses to herbivory (Wal 2006)

as these also stimulate the increase of nutrient-demanding graminoids (Barthelemy

et al. 2014).

32

Fig. 6. The impact of herbivory on the vegetation in Raisduoddar, Čearro and Kilpisjärvi

(from left to right). In Raisduoddar, the mosaic of evergreen and deciduous shrubs with

a thick bryophyte layer was gradually replaced by graminoids with increasing herbivory

intensity (from bottom to top the figures represent light, moderate and heavy grazing

intensity). In Čearro, Betula nana persisted under light and moderate grazing and the

ground layer changed from lichen dominance under light grazing to bryophyte

dominance under moderate grazing. Under heavy grazing, the bryophytes were reduced

and the vegetation was graminoid-dominated. In Kilpisjärvi, the simulated herbivory

replaced the dominant Vaccinium myrtillus by the evergreen Empetrum nigrum ssp.

hermaphroditum.

Fig. 7. The impact of herbivory on ecosystem carbon stocks at the three study sites.

Left: In Raisduoddar, where graminoids replaced mixed shrub vegetation with high

abundance of E. hermaphroditum, soil carbon stocks were not altered by herbivory. In

33

the middle: In Čearro, where B. nana prevailed under light grazing, herbivory induced

grassification of the vegetation and increased soil carbon stocks. Right: In Kilpisjärvi,

herbivory increased the dominance of E. hermaphroditum over V. myrtillus and

decreased soil carbon stocks.

Noteworthy, the mycorrhizal communities associated with the dominant deciduous

shrub species under light grazing differed between Čearro and Kilpisjärvi. In

contrast to the ectomycorrhizal B. nana (Parker et al. 2015), V. myrtillus is

associated with ericoid mycorrhizal fungi (Kasurinen et al. 2001) These associated

mycorrhiza might also explain the differences between sites in soil carbon stocks

under light grazing: there was less soil carbon under the B. nana-dominated

vegetation under both light and moderate grazing in Čearro (1.0 ± 0.1 kg C m-2; 0.9

± 0.1 kg C m-2) compared to V. myrtillus-dominated control plots in Kilpisjärvi (2.3

± 0.3 kg C m-2). This is in line with the observation of a negative correlation

between B. nana and soil carbon stocks (Parker et al. 2015). Moreover, it is peculiar

that the simulated herbivory did not alter the mycorrhizal status that was dominant

in the Kilpisjärvi vegetation.

As the results of carbon stocks do not depict the mechanisms that drive the

changes, I combined these with measurements of CO2 fluxes and microbial activity.

Concurrent with an earlier study (Väisänen et al. 2014) from Raisduoddar, the

present results indicate that ecosystem CO2 sink was stronger under light compared

to heavy grazing (I). These results would give reason to expect more carbon to

accumulate in the ecosystem under light grazing. However, I found no difference

in soil carbon storage between the grazing intensities on a wider area (I) or within

the same plots (II). These results indicate that there are differences between the two

vegetation types either in the carbon accumulation to soil or in alternative pathways

of carbon losses (Lovett et al. 2006). One potential pathway of carbon loss from

the lightly grazed ecosystem could be winter respiration, which is suggested to

correlate with shrub abundance (Sturm et al. 2005, Grogan 2012). In contrast to

these differences in processes in Raisduoddar, in Kilpisjärvi there was no effect of

herbivory on CO2 fluxes (III) or on microbial respiration in either litter or the

organic soil layer (III).

Transient dynamics of herbivore-induced changes

At all study sites, the impact of herbivores on vegetation was strongly dependent

on the duration of herbivory: In Raisduoddar and Čearro, I compared areas with

decadal history (˃ 50 years) of contrasting herbivore pressures. In Čearro, 11 years

34

after the establishment of the reindeer fence, shrub cover was significantly reduced

and there was only a slight increase in graminoids (Oksanen 1978). Only 27 years

later, in 2000, the graminoid biomass was remarkably higher under heavy grazing

when compared to light grazing (Olofsson et al. 2004). In 2014, the graminoid

biomass had continued to increase at both Čearro and Raisduoddar, especially

under moderate grazing in Raisduoddar, where graminoid biomass was 12 times

higher than in 2000. These results indicate that transient dynamics in vegetation

composition may still take place, even after decades of contrasting grazing histories

(Olofsson et al. 2001).

Parallel to the shifts in vegetation, grazing intensity induces substantial

changes in the soil processes and ecosystem carbon fluxes. Already in 1997 and

2000, heavy and moderate grazing by reindeer increased litter decomposition rate,

soil nutrient availability (Olofsson et al. 2001, 2004, Stark et al. 2002) and soil

temperatures in July (Olofsson et al. 2004). In the years 2010–2012, soil

temperatures under heavy grazing were higher than under light grazing on June,

July and August (Väisänen et al. 2014, Stark et al. 2015). Moreover, the

summertime CO2 sink was 70% stronger under light grazing than under heavy

grazing, because respiration was higher under heavy grazing and there were no

differences in ecosystem productivity in response to grazing (Väisänen et al. 2014).

My measurements of the CO2 fluxes in 2014 show that the net CO2 sink was equally

low under moderate and heavy grazing, although ecosystem respiration under

moderate grazing was much lower than under heavy grazing (I). I did not expect

this kind of difference in respiration, because the vegetation communities under

heavy and moderate grazing were highly similar (I). Therefore, I suggest, that

compared to vegetation communities, the belowground processes contributing to

ecosystem respiration might respond to a changing environment with a time lag

(Lamb et al. 2011).

Interestingly, the short-term exclosures under heavy grazing in Raisduoddar

drove the vegetation in the opposite direction as could be anticipated based on the

difference between lightly and heavily grazed areas (II). The higher abundance of

vascular vegetation and graminoids and decreased cover of bryophytes were further

accompanied by increased microbial activity. This indicates that grazer-induced

vegetation changes may not be irreversible, at least in the short-term (Tanentzap et al. 2009), possibly due to the legacy of grazer-induced increase in soil nutrient

availability, which may persist even centuries after grazing (Tømmervik et al. 2010).

35

In Kilpisjärvi, 9 and 13 years after the set-up of the experiment, simulated

herbivory decreased V. myrtillus cover and increased the cover of other shrubs

(Rinnan et al. 2009). Already then, the cover of evergreen shrubs had increased

(20%) but the increase was much stronger (42%) after 16 and 19 years of the

experiment (III). This result indicates that the vegetation community is still

changing even after a decade of continued disturbance.

3.2 Interactive effects of herbivores and climate change on

ecosystem carbon

Aboveground responses to simulated climate change under contrasting

herbivore intensities

At both Raisduoddar and Kilpisjärvi the experimental treatments, warming and

fertilization, increased the abundance of vascular vegetation (II, III) consistent with

previous reports from warming experiments (Elmendorf et al. 2012a, Lu et al. 2013). As hypothesized, the vegetation responses to warming were highly

dependent on herbivory and were in line with my predictions. In the absence of

herbivory (warmed plots under light grazing in Raisduoddar, warmed plots in

Kilpisjärvi), warming especially increased the abundance of deciduous shrubs. The

increased shrubs were mainly B. nana and Salix species in Raisduoddar and V. myrtillus in Kilpisjärvi. These responses to warming are consistent with the general

trend of shrubification that has been observed across the Arctic over the last decades

and within several field experiments (Myers-Smith et al. 2011, 2015). Notably at

both sites, herbivory prevented the warming-induced spread of deciduous shrubs

and instead, promoted the abundance of other species (Post & Pedersen 2008). In

Raisduoddar, the abundance of forbs and grasses was increased by warming under

heavy grazing (II). Within the exclosures under heavy grazing and with the

simulated herbivory in Kilpisjärvi warming increased the abundance of evergreen

shrubs, mainly E. hermaphroditum (II, III) – a response that has been documented

in other warming experiments (Kaarlejärvi et al. 2012, Elmendorf et al. 2012a) and

as a trend in Fennoscandian vegetation (Virtanen et al. 2003). The fertilization in

Raisduoddar increased the abundance of graminoids under all grazing intensities

and decreased the cover and biomass of bryophytes (II). Consequently, fertilization

reduced carbon stocks in bryophytes and increased those in graminoids. Notably,

the treatment-induced changes in vegetation community led to no effects in

36

aboveground carbon stocks in Raisduoddar but in Kilpisjärvi, the carbon stocks

aboveground and belowground were increased by warming, in line with the

majority of warming studies (Lu et al. 2013).

Simulated warming outweighed the impact of herbivory on ecosystem

carbon stocks

Even though experimental warming rarely depicts changes in soil carbon stocks

(Lu et al. 2013), my studies show evidence of a warming-induced decrease in soil

carbon stock in Raisduoddar (II) and a warming-induced increase in soil carbon

stock in Kilpisjärvi (III). These contrasting responses to warming took place despite

that the experimental sites in Kilpisjärvi and Raisduoddar are geographically

relatively close to one another (~ 60 km). At both sites, warming concurrently

increased the rates of gross ecosystem productivity and ecosystem respiration

(Väisänen et al. 2014, III), microbial respiration (Stark et al. unpublished, III), the

potential activity of microbial extracellular enzymes (II, Stark et al. unpublished)

and microbial community composition (II, Rinnan et al. 2009). Warming also

decreased soil moisture (II, III), which is often linked to enhanced microbial

activity (Davidson & Janssens 2006). The effect of warming on soil microbial

activities and ecosystem carbon fluxes were thus similar at both sites. Yet, the

outcome on soil carbon stocks was highly different. These contrasting results

indicate that in Kilpisjärvi the positive effect of warming on carbon uptake

outpaced the negative effects of higher plant and microbial respiration, whereas the

opposite was true in Raisduoddar and the carbon losses outpaced the carbon gains.

Common to both experiments was, however, that herbivory (either simulated,

the short-term exclosure or the decadal difference in grazing intensity) did not alter

the direction of warming-induced change on soil carbon. This occurred despite the

contrasting, herbivory-depended, responses of vegetation communities to warming.

The result contrasted my hypothesis, where I expected the effects of warming on

ecosystem carbon stocks to be coupled with vegetation communities and the

changes in it. I suggest that this decoupling of vegetation and belowground

responses may be to some extent due to the relatively short experiment time in

Raisduoddar (5 years) or the slow feedback of vegetation to simulated herbivory in

Kilpisjärvi (significant increases in evergreens within five years of sampling). Also

Bradford et al. (2016) have suggested that the short-term feedbacks to experimental

warming are likely different from those in the long-term. As an example, several

studies reported carbon losses after experimental warming in a tussock tundra in

37

Alaska (Hobbie & Chapin 1998, Shaver et al. 2006, Oberbauer et al. 2007) but

when ecosystem carbon stocks were assessed 19 years after the establishment of

the experiment, no warming-induced changes in total soil carbon were found (Sistla

et al. 2013). In experimental time-scales, the Kilpisjärvi experiment is

exceptionally long and the soil community is likely to have adapted to the continued

warming within the 19 years of the experiment. However, I suggest that the soil

communities might not have yet adapted to the altered litter input under simulated

herbivory as the vegetation response was relatively recent.

Although these results support the view that warmer temperatures enhance soil

microbial activity for carbon cycling (Davidson & Janssens 2006) and alter

microbial community composition (Karhu et al. 2014), this data only gives a coarse

estimation of the carbon stocks changes and some processed behind it. Notably,

there may have been community-depended differences in other processes, such as

warming-induced changes in the production of root exudates (Iversen et al. 2015),

the carbon-transfer to mycorrhizal networks (Deslippe et al. 2012), the leaching of

carbon (Lu et al. 2013) and the distribution of roots (Sistla et al. 2013). These

studies also lack thorough information on the carbon fluxes in the winter-time,

which constitute a major part of the net carbon exchange in the tundra (Nobrega &

Grogan 2007, Grogan 2012).

Ecosystem responses to enhanced nutrient-availability under contrasting

herbivory intensities

In the Raisduoddar study, experimental fertilization exerted drastic effects on

vegetation: vascular vegetation and graminoids were increased and the biomass of

bryophytes was decreased (II). Concurrently, fertilization altered the potential

activities of the microbial extracellular enzymes β-glucosidase and N-acetyl-

glucosaminidase indicating increased rates of soil carbon and microbial turnover

(II, Allison et al. 2010, Beier & Bertilsson 2013). Notably, the impact of

fertilization on vegetation abundance and microbial activities was stronger than that

of warming. Yet, fertilization at its own had negligible effects on the carbon stocks

belowground. These results indicate that predictions on the basis of individual

processes may not be consistent with outcomes because several mechanisms may

act at the same time, possibly in opposite directions (Lu et al. 2013).

I suggest that many of the fertilization-induced changes in Raisduoddar (II)

derived from the increased litter input on fertilized plots as fertilization increased

the abundance of graminoids and decreased bryophyte cover. The rapid depletion

38

of the bryophyte layer could depict toxicity of fertilization on bryophytes (Bubier

et al. 2007) and result in higher litter loads that are rich in carbohydrates

(Vancampenhout et al. 2009). The higher litter input from graminoids and

bryophytes could explain the increased microbial carbon acquisition (as indicated

by the potential activity of β-glucosidase), the higher fungal:bacterial ratio and

higher microbial turnover rate (potential N-acetyl-glucosaminidase activity)

because fungi benefit from litter additions (Rousk & Frey 2015). This steady carbon

input to study plots could to some level explain why we did not detect a

fertilization-induced decrease in soil carbon stocks (Mack et al. 2004) – not even

under light grazing, where the vegetation was dominated by deciduous and

evergreen shrubs.

When fertilization was combined with warming, the impacts on vegetation and

microbial activities were amplified (II). Importantly, fertilization negated the

warming-induced soil carbon loss under the long-term heavy and light grazing

pressures, but within the short-term exclosures organic soil carbon was lost even in

the combined warming and fertilization treatment (II). The highly similar response

under the contrasting long-term herbivory intensities shows that these systems were

equally resistant to abiotic changes despite the herbivory-induced differences in

nutrient cycling rates (Olofsson et al. 2004, Stark & Väisänen 2014), soil

temperatures (Olofsson et al. 2004, Väisänen et al. 2014, I) and microbial

community compositions (Männistö et al. 2016). This result suggests that, under

steady herbivore intensities, the carbon–climate feedback in the tundra could be

diminished if nitrogen availability is simultaneously enhanced (but see Mack et al. 2004). Therefore, current and future extreme events affecting nitrogen availability

could have consequences on ecosystem functioning.

Notably, fertilization did not negate warming-induced carbon losses within the

short-term herbivory exclosures. The result highly contrasts previous studies where

the exclusion of herbivores has intensified fertilization-induced increases in

ecosystem CO2 sink or carbon sequestration (Smith et al. 2015). The contrasting

responses indicate that there may be regional differences in the responses and call

for improved understanding of the mechanisms and patterns behind the interactions

between herbivory and altered abiotic factors. Moreover, I suggest that biotic

changes may interact with abiotic changes and intensify ecosystem responses to

these.

39

Implications on the interactive effects of herbivores and abiotic changes

on tundra carbon storage

The results from the two experimental warming studies indicate that, within short

time-scales, the direct effect of warming may deluge the impact of the dominant

vegetation and drive the changes in soil carbon stocks. I suggest that, given more

time, the herbivory-driven change in vegetation community would still likely

influence soil carbon storage at both sites – even in the presence of warming. The

control of plant functional traits over soil carbon sequestration has been reported

within biomes worldwide (Deyn et al. 2008) which indicates that the relationship

is independent of the abiotic environment. Therefore, I propose that with enough

time for communities to adapt and processes to reach a steady-state level, plant

functional traits will continue to affect carbon stocks. The future changes in soil

carbon storage could be, to some level, predicted based on the traits of the

vegetation that increased in response to the treatments – as I predicted in my

hypotheses. At both sites, such predictions would indicate a positive effect of

herbivores on ecosystem carbon sequestration, because both graminoids (increased

in Raisduoddar under heavy grazing and experimental warming) and evergreen

shrubs (increased in Kilpisjärvi under simulated herbivory and experimental

warming) may produce litter of higher recalcitrancy compared to deciduous shrubs

(increased by warming under light / no herbivory at both sites; Deyn et al. 2008,

Freschet et al. 2013).

My results also showed that short-term cessation of grazing increased

graminoids and decreased the bryophyte layer. If the cessation of grazing would

decrease graminoids’ resource allocation to roots (Gao et al. 2007, Hafner et al. 2012), the ceased grazing could potentially decrease soil carbon sequestration.

However, within the exclosures, warming favored the establishment of evergreen

shrubs and bryophytes, which produce recalcitrant litter and may decrease soil

temperatures (Chapin et al. 1996), potentially indicating that warming could

increase ecosystem carbon sequestration after the cessation of grazing (Cahoon et al. 2012b). Notably, we evidenced an opposite pattern when herbivore exclusion

and warming were accompanied with fertilization: these again favoured graminoid

abundance and the soil carbon storage was decreased. I could not depict whether

this was linked to the responses of specific plant species or microbial responses.

We also found that fertilization negated the warming-induced carbon loss under

both long-term grazing intensities. I suggest that this was not linked to vegetative

responses as fertilization tentatively decreased root biomass under heavy grazing

40

(II), which would indicate a negative effect of fertilization on soil carbon

sequestration. It is possible that the observed negation of carbon loss was rather

linked to microbial activities and derived from a weaker capacity of soil microbial

community to decompose recalcitrant carbon compounds (Fontaine et al. 2003) as

suggested by a soil incubation experiment in the same site (Männistö et al. 2016).

The predictions based on functional group changes in both Raisduoddar and

Kilpisjärvi indicate that herbivory could increase soil carbon sequestration in the

long-term. Thus, instead of protecting areas from herbivory, the present herbivory

patterns – and even an introduction of higher herbivore intensities – could maintain

the highest rates of carbon uptake – at least on certain areas. To some level this

generalization could be influenced by the experimental design of simulated

herbivory in Kilpisjärvi or the site-specific grazing pattern in Raisduoddar, where

the high herbivory intensity is limited to only a few weeks in the growing season.

Therefore, results from both studies rather depict how short-term pulses of high

herbivore intensity impact ecosystem processes and carbon storage. An improved

understanding of the conditions under which such pulses of herbivory promote soil

carbon sequestration would be highly useful for land-use planning and the

management of grazers. Recommendations on the basis of further studies could be

particularly applicable in Fennoscandia where reindeer husbandry is already

considered to be an important means of land-use (Forbes & Kumpula 2009).

Moreover, the results from both studies indicate that predictions of tundra carbon

balance might be improved if herbivory was included in the climate change models.

This could also improve estimations of the tundra carbon–climate feedback.

41

4 Conclusions

In this thesis, I describe several means by which herbivory and climate change can

affect ecosystem carbon storage. I show that in the current climate, the subarctic

ecosystem continued to accumulate carbon. This was seen as an increase in

ecosystem carbon stocks over the past 14 years and the observed negative growing

seasonal net ecosystem exchange in the study sites. Concurrent with my hypothesis

and the theory that plant functional traits control carbon sequestration, my results

demonstrate strong herbivore control over ecosystem carbon storage. When

herbivory increased the abundance of evergreen over deciduous shrubs, the

aboveground carbon stocks were increased and those belowground were decreased

resulting in no change in total ecosystem carbon. When herbivory increased the

abundance of graminoids over deciduous or evergreen shrubs, the aboveground

carbon stocks were decreased and the stocks belowground were either increased or

did not change. These findings indicate that the net effect of herbivory on

ecosystem carbon stocks possibly depends on the traits of the species that are

replaced by herbivory. Consequently, I propose that the graminoid-dominated

tundra vegetation lies, in its ability to sequester carbon, between the deciduous

shrub tundra (low carbon sequestration) and the evergreen tundra heath (high

carbon sequestration). The result that is not in line with this suggestion – nor with

the prediction based on plant functional traits – was that through simulated

herbivory the replacement of deciduous shrubs by evergreens decreased soil carbon

storage. I explain the discrepancy by the short time that passed since the vegetation

change and by the yearly removal of plant biomass, and thus carbon, from the study

plots.

To investigate the interactive effects of climate-change and herbivory, I applied

warming or warming and fertilization under contrasting levels of herbivory. I found

strong herbivory control in the warming-induced changes of vegetation and the

plant functional traits but no effect of herbivory on the vegetation responses to

fertilization. Yet, the sole effects of warming were the same under contrasting

herbivore intensities and, instead, the responses to combined warming and

fertilization varied with herbivory. These results contrasted my hypothesis and

suggest that the short-term responses of ecosystem carbon storage to both warming

and fertilization may not be coupled with herbivory-induced shifts in the dominant

plant functional traits.

I showed that the response of soil carbon stocks to experimental warming

depended on the study site, and warming either decreased or increased soil carbon

42

stocks. Nevertheless, at both sites, warming was paralleled with increased

vegetation abundance, gross ecosystem production, ecosystem respiration and soil

microbial activity and decreased soil moisture. These responses indicate that

individual processes may respond fairly similarly to warming, but the outcome of

these can vary from positive to negative. In line with this, fertilization exerted

strong control on the vegetation and microbial activities, but, when applied alone,

it did not alter soil carbon storage. However, when fertilization was applied with

warming, it negated the warming-induced carbon loss under both long-term

herbivory intensities, but not when herbivory was abruptly removed. This result

indicates that sudden biotic changes may interact with abiotic changes and possibly

result in a more unpredictable outcome.

Even though results from both of the experimental studies gave no support to

my hypothesis of plant functional traits governing ecosystem functions in a warmer

climate, I do believe that in the long-term different vegetation communities will

vary in their carbon sequestration rates even when temperatures have increased. I

base this assumption on observations of vegetation control of soil carbon storage

within tundra and other biomes worldwide. Therefore, experimental studies, such

as the ones in this thesis, shed light on the conditions under which certain plant

species and functional traits are favored. Importantly, my studies demonstrate that

the vegetation responses to abiotic changes depended on the ambient herbivory

intensity. As the growth forms favoured by higher grazing intensities, namely

graminoids and evergreen shrubs, generally maintain higher rates of carbon

sequestration rates than deciduous shrubs, this gives insight that grazing could

actually be beneficial for the carbon cycling in the long-term.

Taken together, I suggest that herbivory control over vegetation dynamics has

a highly overlooked role in shaping ecosystem processes and carbon storage. The

wide distribution of herbivores around the globe and particularly in tundra areas

gives reason to suggest that the regional effects of herbivores might sum up to be

of global importance. An improved understanding of the role of herbivory could

improve estimations of global carbon–climate feedbacks and could be helpful in

planning land-use with the aim to increase carbon sequestration or prevent carbon

losses.

43

List of references

Allison SD, Wallenstein MD & Bradford MA (2010) Soil-carbon response to warming dependent on microbial physiology. Nature Geoscience 3: 336–340.

Averill C, Turner BL & Finzi AC (2014) Mycorrhiza-mediated competition between plants and decomposers drives soil carbon storage. Nature 505: 543–545.

Bardgett RD & Wardle D (2003) Herbivore-mediated linkages between aboveground and belowground communities. Ecology 84: 2258–2268.

Bardgett RD, Mommer L & De Vries FT (2014) Going underground: root traits as drivers of ecosystem processes. Trends in Ecol & Evol 29: 692–699.

Barnosky AD, Lindsey EL, Villavicencio NA, Bostelmann JE, Hadly EA, Wanket J, Marshall CR (2016) Variable impact of Late Quaternary megafauna extinction in causing ecological state shifts in North and South America. PNAS 113: 856–861.

Barrio IC Bueno CG, Gartzia M, Soininen EM, Christie KS, Speed JDM, Ravolainen VT, Forbes BC, Gauthier G, Horstkotte T, Hoset KS, Høye TT, Jónsdóttir IS, Lévesque E, Mörsdorf MA, Olofsson J, Wookey PA & Hik DS (2016) Biotic interactions mediate patterns of herbivore diversity in the Arctic. Glob Ecol Biogeogr 25: 1108–1118.

Barthelemy H, Stark S & Olofsson J (2014) Strong responses of subarctic plant communities to long-term reindeer feces manipulation. Ecosystems 18: 740–751.

Beier S & Bertilsson S (2013) Bacterial chitin degradation-mechanisms and ecophysiological strategies. Front Microbiol 4: 149.

Beisner BE, Haydon DT & Cuddington K (2003) Alternative stable states in ecology. Front Ecol Environ 1: 376–382.

Bernes C, Bråthen KA, Forbes BC, Speed JDM & Moen J (2015) What are the impacts of reindeer/caribou (Rangifer tarandus L.) on arctic and alpine vegetation? A systematic review. Environ Evid 4: 1–26.

Bokhorst S, Huiskes A, Coney P, Sinclair BJ, Lebouvier M, Van de Vijver B & Wall DH (2011) Microclimate impacts of passive warming methods in Antarctica: implications for climate change studies. Polar Biol 34: 1421–1435.

Bonfils CJW, Phillips TJ, Lawrence DM, Cameron-Smith P, Riley WJ & Subin ZM (2012) On the influence of shrub height and expansion on northern high latitude climate. Environ Res Lett 7: 9.

Bradford MA, Wieder WR, Bonan GB, Fierer N, Raymond PA & Crowther TW (2016) Managing uncertainty in soil carbon feedbacks to climate change. Nat Clim Chang 6: 751–758.

Bubier JL, Moore TR & Bledzki LA (2007) Effects of nutrient addition on vegetation and carbon cycling in an ombrotrophic bog. Glob Chang Biol 13: 1168–1186.

Cahoon SMP, Sullivan PF, Shaver GR, Welker JM & Post E (2012a) Interactions among shrub cover and the soil microclimate may determine future Arctic carbon budgets. Ecol Lett 15: 1415–1422.

Cahoon SMP, Sullivan PF, Post E & Welker JM (2012b) Large herbivores limit CO2 uptake and suppress carbon cycle responses to warming in west Greenland. Glob Chang Biol 18: 469–479.

44

Chapin FS III, Bret-Harte MS, Hobbie SE & Zhong HL (1996) Plant functional types as predictors of transient responses of arctic vegetation to global change. J Veg Sci 7: 347–358.

Clemmensen K, Bahr A, Ovaskainen O, Dahlberg A, Ekblad A, Wallander H, Stenlid J, Finlay R, Wardle D & Lindahl B (2013) Roots and associated fungi drive long-term carbon sequestration in boreal forest. Science 339: 1615–1618.

Cornelissen JHC, Lang SI, Soudzilovskaia NA & During HJ (2007). Comparative cryptogam ecology: a review of bryophyte and lichen traits that drive biogeochemistry. Ann Bot 99: 987–1001.

Cornwell WK, Cornelissen JHC & Amatangelo K (2008) Plant species traits are the predominant control on litter decomposition rates within biomes worldwide. Ecol Lett 11: 1065–1071.

Crowther T, Todd-Brown KEO, Rowe CW, Wieder WR, Carey JC, Machmuller MB, Snoek BL, Fang S, Zhou G, Allison SD, Blair JM, Bridgham SD, Burton AJ, Carrillo Y, Reich PB, Clark JS, Classen AT, Dijkstra FA, Elberling B, Emmett BA, Estiarte M, Frey SD, Guo J, Harte J, Jiang L, Johnson BR, Kröel-Dulay G, Larsen KS, Laudon H, Lavallee JM, Luo Y, Lupascu M, Ma LN, Marhan S, Michelsen A, Mohan J, Niu S, Pendall E, Peñuelas J, Pfeifer-Meister L, Poll C, Reinsch S, Reynolds LL, Schmidt I.K, Sistla S, Sokol NW, Templer PH, Treseder KK, Welker JM & Bradford MA (2016) Quantifying global soil C losses in response to warming. Nature 104: 104–108.

Dallimer M, Tinch D, Acs S, Hanley N, Southall HR, Gaston KJ & Armsworth PR (2009) 100 years of change: examining agricultural trends, habitat change and stakeholder perceptions through the 20th century. J Appl Ecol 46: 334–343.

Davidsson EA & Janssens IA (2006) Temperature sensitivity of soil carbon decomposition and feedbacks to climate change. Nature 440: 165–173.

Deslippe JR, Hartmann M, Simard SW & Mohn WW (2012) Long-term warming alters the composition of Arctic soil microbial communities. FEMS Microbiol Ecol 82: 303–315.

Deyn GB de, Cornelissen JHC & Bardgett RD (2008) Plant functional traits and soil carbon sequestration in contrasting biomes. Ecol Lett 11: 516–531.

Dorrepaal E (2007) Are plant growth-form-based classifications useful in predicting northern ecosystem carbon cycling feedbacks to climate change? J Ecol 95: 1167–1180.

Elliott TL & Henry GHR (2011) Effects of simulated grazing in ungrazed wet sedge tundra in the High Arctic. Arct Antarct Alp Res 43: 198–206.

Elmendorf S, Henry G, Hollister R, Björk R, Bjorkman A, Callaghan T, Collier L, Cooper E, Cornelissen J, Day T, Fosaa A, Gould W, Grétarsdóttir J, Harte J, Hermanutz L, Hik D, Hofgaard A, Jarrad F, Jónsdóttir I, Keuper F, Klanderud K, Klein J, Koh S, Kudo G, Lang S, Loewen V, May J, Mercado J, Michelsen A, Molau U, Myers‐Smith I, Oberbauer S, Pieper S, Post, E, Rixen C, Robinson C, Schmidt N, Shaver G, Stenström A, Tolvanen A, Totland Ø, Troxler T, Wahren C, Webber P, Welker J & Wookey P (2012a) Global assessment of experimental climate warming on tundra vegetation: Heterogeneity over space and time. Ecol Lett 15: 164–175.

45

Elmendorf S, Henry G, Hollister R, Björk R, Boulanger-Lapointe N, Cooper E, Cornelissen J, Day T, Dorrepaal E, Elumeeva T, Gill M, Gould W, Harte J, Hik D, Hofgaard A, Johnson D, Johnstone J, Jónsdóttir I, Jorgenson J, Klanderud K, Klein J, Koh S, Kudo G, Lara M, Lévesque E, Magnússon B, May J, Mercado-Díaz J, Michelsen A, Molau U, Myers-Smith I, Oberbauer S, Onipchenko V, Rixen C, Schmidt N, Shaver G, Spasojevic M, Þórhallsdóttir Þ, Tolvanen A, Troxler T, Tweedie C, Villareal S, Wahren C-H, Walker X, Webber P, Welker J & Wipf S (2012b) Plot-scale evidence of tundra vegetation change and links to recent summer warming. Nat Clim Chang 1465: 453–457.

Falk J, Schmidt N, Christensen T & Ström L (2015) Large herbivore grazing affects the vegetation structure and greenhouse gas balance in a high arctic mire. Environ Res Lett 10: 045001.

Fontaine S, Mariotti A & Abbadie L (2003) The priming effect of organic matter: a question of microbial competition? Soil Biol Biochem 35: 837e843.

Forbes BC & Kumpula T (2009) The ecological role and geography of reindeer (Rangifer tarandus) in northern Eurasia. Geograph Compass 3: 1356–1380.

Freschet GT, Cornwell WK, Wardle DA, Elomeeva TG, Liu W, Jackson BG, Onipchenko VG, Soudzilovskaia NA, Tao J & Cornelissen JHC (2013) Linking litter decomposition of above- and below-ground organs to plant–soil feedbacks worldwide. J Ecol 101: 943–952.

Gao YH, Luo P, Wu N, Chen H & Wang GX (2007) Grazing intensity impacts on carbon sequestration in an alpine meadow on the eastern Tibetan Plateau. Res J Agric & Biol Sci 3: 642–647.

Gornall J, Jónsdóttir I, Woodin S & Wal R van der (2007) Arctic bryophytes govern below-ground environment and ecosystem processes. Oecol 153: 931–941.

Gornall J, Woodin S. Jónsdóttir I & Wal R van der (2009) Herbivore impacts to the bryophyte layer determine tundra ecosystem response to grazing and warming. Oecol 161: 747–758.

Grogan P (2012) Cold season respiration across a low arctic landscape: The influence of vegetation-type and interannual climatic variation. Arct Antarc Alp Res 44: 446–456.

Hafner S, Unteregelsbacher S, Seeber E, Lena B, Xu X, Li X, Guggenberger G, Miehe G & Kuzyakov Y (2012) Effect of grazing on carbon stocks and assimilate partitioning in a Tibetan montane pasture revealed by 13-CO2 pulse labeling. Glob Chang Biol 18: 528–538.

Hartley I, Garnett M, Sommerkorn M, Hopkins D, Fletcher B, Sloan V, Phoenix G & Wookey P (2012) A potential loss of carbon associated with greater plant growth in the European Arctic. Nat Clim Chang 2: 875–879.

Hobbie S & Chapin FS III (1998) The response of tundra plant biomass, aboveground production, nitrogen, and CO2 flux to experimental warming. Ecology 79: 1526–1544.

Hollister RD & Webber PJ (2000) Biotic validation of small open-top chambers in a tundra ecosystem. Glob Chang Biol 6: 835–842.

46

Ihrmark K, Bödeker IT, Cruz-Martinez K, Friberg H, Kubartova A, Schenck J, Strid Y, Stenlid J, Brandström-Durling M, Clemmensen KE & Lindahl BD (2012) New primers to amplify the fungal ITS2 region - evaluation by 454-sequencing of artificial and natural communities. FEMS Microbiol Ecol 82: 666–677.

IPCC (2014) Summary for Policymakers. In: Field CB, Barros VR, Dokken D J, Mach KJ, Mastrandrea MD, Bilir TE, Chatterjee M, Ebi KL, Estrada YO, Genova RC, Girma B, Kissel ES, Levy AN, MacCracken S, Mastrandrea PR & White LL (eds) Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part A: Global and Sectoral Aspects. Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge and New York NY, Cambridge University Press: 1–32.

Iversen CM, Sloan VL, Sullivan PF, Euskirchen ES, McGuire AD, Norby RJ, Walker AP, Warren JM & Wullschleger SD (2015) The unseen iceberg: Plant roots in arctic tundra. New Phytol 205: 34–58.

Jonasson S (1988) Evaluation of the point intercept method for the estimation of plant biomass. Oikos 52: 101–106.

Kaarlejärvi E, Baxter R, Hofgaard A, Hytteborn H, Khitun O, Molau U, Sjögersten S, Wookey P & Olofsson J (2012) Effects of warming on shrub abundance and chemistry drive ecosystem-level changes in a forest-tundra ecotone. Ecosystems 15: 1219–1233.

Karhu K, Auffret MD, Dungait JA, Hopkins DW, Prosser JI, Singh BK, Subke J-A, Wookey PA, Agren GI, Sebastia M-T, Gouriveau F, Bergkvist G, Meir P, Nottingham AT, Salinas N & Hartley IP (2014) Temperature sensitivity of soil respiration rates enhanced by microbial community response. Nature 513: 81–84.

Kasurinen A, Holopainen T, & Anttonen S (2001) Mycorrhizal colonisation of highbush blueberry and its native relatives in central Finland. Agric Food Sci Finl 10: 113–119.

Knapp R (1984) Sampling methods and taxon analysis in vegetation science. Handbook of Vegetation Science 1. Part 4. Hague, Junk.

Köster K, Berninger F, Köster E & Pumpanen J (2015) Influences of reindeer grazing on above- and belowground biomass and soil carbon dynamics. Arct Antarct Alp Res 47: 495–503.

Lamb E, Han S, Lanoil B, Henry G, Brummel M, Banerjee S & Siciliano S. (2011) A High Arctic soil ecosystem resists long-term environmental manipulations. Glob Chang Biol 17: 3187–3194.

Lovett G, Cole J & Pace M (2006) Is net ecosystem production equal to ecosystem carbon accumulation? Ecosystems 9: 152–155.

Lu M Zhou X, Yang Q, Li H, Luo Y, Fang C, Chen J, Yang X & Li B (2013) Responses of ecosystem carbon cycle to experimental warming: a meta-analysis. Ecology 94: 726–738.

Mack MC, Schuur EAG, Bret-Harte MS, Shaver GR & Chapin FS (2004) Ecosystem carbon storage in Arctic tundra reduced by long-term nutrient fertilization. Nature 43: 440–443.

Männistö M, Ganzert L, Tiirola M, Häggblom M & Stark S (2016) Do shifts in life strategies explain microbial community responses to increasing nitrogen in tundra soil? Soil Biol Biochem 96: 216–228.

47

Marion GM, Henry GHR, Freckman DW, Johnstone J, Jones G, Jones MH, Lévesque E, Molau U, Mølgaard P, Parsons AN, Svoboda J & Virginia RA (1997) Open-top designs for manipulating field temperature in high latitude ecosystems. Glob Chang Biol 3: 20–31.

Martinsen V, Mulder J, Austrheim G & Mysterud A (2011) Carbon storage in low-alpine grassland soils: effects of different grazing intensities of sheep. Eur J Soil Sci 62: 822–833.

Marushchak ME, Kiepe I, Biasi C, Elsakov V, Friborg T, Johansson T, Soegaard H, Virtanen T & Martikainen PJ (2013) Carbon dioxide balance of subarctic tundra from plot to regional scales. Biogeosciences 10: 437–452.

Metcalfe DB & Olofsson J (2015) Distinct impacts of different mammalian herbivore assemblages on arctic tundra CO2 exchange during the peak of the growing season. Oikos 124: 1632–1638.

Myers-Smith IH, Forbes B, Wilmking M, Hallinger M, Lantz T, Blok D, Tape K, Macias-Fauria M, Sass-Klaassen U, Lévesque E, Boudreau S, Ropars P, Hermanutz L, Trant A, Collier LS, Weijers S, Rozema J, Rayback S, Schmidt N, Schaepman-Strub G, Wipf S, Rixen C, Ménard C, Venn S, Goetz S, Andreu-Hayles L, Elmendorf S, Ravolainen V, Welker J, Grogan P, Epstein H & Hik D (2011) Shrub expansion in tundra ecosystems: dynamics, impacts and research priorities. Environ Res Lett 6: 045509.

Myers-Smith IH, Elmendorf S, Beck P, Wilmking M, Hallinger M, Blok D, Tape K, Rayback S, Macias-Fauria M, Forbes B, Speed J, Boulanger-Lapointe N, Rixen C, Lévesque E, Schmidt N, Baittinger C, Trant A, Hermanutz L, Collier L, Dawes M, Lantz T, Weijers S, Jørgensen R, Buchwal A, Buras A, Naito A, Ravolainen V, Schaepman-Strub G, Wheeler J Wipf S, Guay K, Hik D & Vellend M (2015) Climate sensitivity of shrub growth across the tundra biome. Nat Clim Chang 5: 887–891.

Nobrega S & Grogan P (2007) Deeper snow enhances winter respiration from both plant associated and bulk soil carbon pools in birch hummock tundra. Ecosystems 10: 419–431.

Nobrega S & Grogan P (2008) Landscape and ecosystem-level controls on net carbon dioxide exchange along a natural moisture gradient in Canadian low Arctic tundra. Ecosystems 11: 377–396.

Nowak RM (1999) Walker's Mammals of the World. Baltimore, John Hopkins University Press.

Oberbauer S, Tweedie C & Welker J (2007) Tundra CO2 fluxes in response to experimental warming across latitudinal and moisture gradients. Ecol Monogr 77: 221–238.

Oksanen J, Blanchet GF, Kindt R, Simpson GL, Solymos P, Stevens HMH & Wagner HH (2015) vegan: Community Ecology Package. Ordination methods, diversity analysis and other functions for community and vegetation ecologists. Version 23-1 URL: https://CRANR-projectorg/package=vegan.

Oksanen L (1978) Lichen grounds of Finnmarksvidda, northern Norway, in relation to summer- and winter grazing by reindeer. Reports of the Kevo Subarctic Research Station 14: 64–71.

48

Olofsson J, Kitti H, Rautiainen P Stark S & Oksanen L (2001) Effects of summer grazing by reindeer on composition of vegetation, productivity and nitrogen cycling. Ecography 24: 13–24.

Olofsson J, Stark S & Oksanen L (2004) Reindeer influence on ecosystem processes in the tundra. Oikos 105: 386–396.

Parker TC, Subke J-A & Wookey PA (2015) Rapid carbon turnover beneath shrub and tree vegetation is associated with low soil carbon stocks at a subarctic treeline. Glob Chang Biol 21: 2070–2081.

Pinheiro J, Bates D, DebRoy S & Sarkar D (2014) R Core Team (2014) nlme: linear and nonlinear mixed effects models. R package version 3.1-117. URL: http://CRAN. R-project. org/package= nlme.

Post E & Pedersen C (2008) Opposing plant community responses to warming with and without herbivores. PNAS 105: 12353–12358.

Rasse DP, Rumpel C & Dignac M-F (2005) Is soil carbon mostly root carbon? Mechanisms for a specific stabilization. Plant Soil 269: 341–356.

Rinnan R, Stark S & Tolvanen A (2009) Responses of vegetation and soil microbial communities to warming and simulated herbivory in a subarctic heath. J Ecol 97: 788–800.

Rousk J & Frey SD (2015) Revisiting the hypothesis that fungal-to-bacterial dominance characterizes turnover of soil organic matter and nutrients. Ecol Monogr 85: 457–472.

Schmitz O, Raymond P, Estes J, Kurz W, Holtgrieve G, Ritchie M, Schindler D, Spivak A, Wilson R, Bradford M, Christensen V, Deegan L, Smetacek V, Vanni M & Wilmers C (2014) Animating the carbon cycle. Ecosystems 17: 344–359.

Scholes RJ & Archer SR (1997) Tree-grass interactions in savannas. Annu Rev Ecol Syst 28: 517–544.

Schuur EAG, McGuire AD, Schädel C, Grosse G, Harden JW, Hayes DJ, Hugelius G, Koven CD, Kuhry P, Lawrence DM, Natali SM, Olefeldt D, Romanovsky VE, Schaefer K, Turetsky MR, Treat CC & Vonk JE (2015) Climate change and the permafrost carbon feedback. Nature 520: 171–179.

Shaver GR, Giblin AE, Nadelhoffer KJ, Thieler KK, Downs MR, Laundre JA & Rastetter EB (2006) Carbon turnover in Alaskan tundra soils: effects of organic matter quality, temperature, moisture and fertilizer. J Ecol 94: 740–753.

Shaver GR, Street LE, Rastetter EB, van Wijk MT & Williams M (2007) Functional convergence in regulation of net CO2 flux in heterogeneous tundra landscapes in Alaska and Sweden. J Ecol 95: 802–817.

Sinsabaugh RL (2010) Phenol oxidase, peroxidase and organic matter dynamics of soil. Soil Biol. Biochem. 42: 391–404.

Sistla, SA, Moore JC, Simpson RT, Gough L, Shaver GR & Schimel JP (2013) Long-term warming restructures Arctic tundra without changing net soil carbon storage. Nature 497: 615–618.

Sjögersten S, Wal R van der & Woodin SJ (2012) Impacts of Grazing and Climate Warming on C Pools and Decomposition Rates in Arctic Environments. Ecosystems 15: 349–362.

49

Smith SW, Johnson D, Quin SL, Munro K, Pakeman RJ, Wal R van der & Woodin SJ (2015) Combination of herbivore removal and nitrogen deposition increases upland carbon storage. Glob Chang Biol 21: 3036–3048.

Speed J, Woodin SJ, Tømmervik H & Wal R van der (2010) Extrapolating herbivore-induced carbon loss across an arctic landscape. Polar Biol 33: 789–797.

Stark S & Väisänen M (2014) Insensitivity of soil microbial activity to temporal variation in soil N in subarctic tundra: evidence from responses to large migratory grazers. Ecosystems 17: 906–917.

Stark S, Strommer R & Tuomi J (2002) Reindeer grazing and soil microbial processes in two suboceanic and two subcontinental tundra heaths. Oikos 97: 69–78.

Stark S, Kytöviita M-M, Männistö MK & Neumann AB (2008) Soil microbial and microfaunal communities and organic matter quality in reindeer winter and summer ranges in Finnish subarctic mountain birch forests. Appl Soil Ecol 40: 456–464.

Stark S, Männistö MK & Smolander A (2010) Multiple effects of reindeer grazing on the soil processes in nutrient-poor northern boreal forests. Soil Biol & Biochem 42: 2068–2077.

Stark S, Männistö M, Ganzert L, Tiirola M & Häggblom M (2015) Grazing influences temperature adaptation of soil microbial community in tundra soil. Soil Biol Biochem 84: 147–157.

Sturm M, Schimel J, Michaelson G, Welker JM, Oberbauer SF, Liston GE, Fahnestock J & Romanovsky VE (2005) Winter biological processes could help convert Arctic tundra to shrubland. BioScience 55: 17–26.

Susiluoto S, Rasilo T, Pumpanen J & Berninger F (2008) Effects of grazing on the vegetation structure and carbon dioxide exchange of a fennoscandian fell ecosystem. Arct Antarct Alp Res 40: 422–431.

Tanentzap AJ & Coomes DA (2012) Carbon storage in terrestrial ecosystems: do browsing and grazing herbivores matter? Biol Revi 87: 72–95.

Tanentzap AJ, Burrows LE, Lee WG, Nugent G, Maxwell JM & Coomes DA (2009) Landscape-level vegetation recovery from herbivory: progress after four decades of invasive red deer control. J Appl Ecol 46: 1064–1072.

Tarnocai C, Canadell JG, Schuur EAG, Kuhry P, Mazhitova G & Zimov S (2009) Soil organic carbon pools in the northern circumpolar permafrost region. Glob Biogeochem Cycl 23: GB2023.

The Plant List (2013). Version 1.1. Published on the Internet; http://www.theplantlist.org/ (accessed 1st January 2017).

Tømmervik H, Dunfjeld S, Olsson GA & Nilsen MØ (2010) Detection of ancient reindeer pens, cultural remains and anthropogenic influenced vegetation in Byrkije (Børgefjell) mountains, Fennoscandia. Landsc. Urban Plan 98: 56–71.

Väisänen M, Ylänne H, Kaarlejärvi E, Sjögersten S, Olofsson J, Crout N & Stark S (2014) Consequences of warming on tundra carbon balance determined by reindeer grazing history. Nat Clim Chang 4: 384–388.

50

Vancampenhout K, Wouters K, De Vos B, Buurman P, Swennen R, & Deckers J (2009) Differences in chemical composition of soil organic matter in natural ecosystems from different climatic regions – a pyrolysis-GC/MS study. Soil Biol Biochem 41: 568–579.

Väre H, Ohtonen R & Oksanen J (1995) Effects of reindeer grazing on understorey vegetation in dry Pinus sylvestris forests. J Veg Sci 6: 523–530.

Virtanen R, Henttonen H & Laine KM (1997) Lemming grazing and the structure of snowbed plant community–a long-term experiment at Kilpisjärvi, in Finnish Lapland. Oikos 79: 155–166.

Virtanen R, Parviainen J & Henttonen H (2002) Winter grazing by the Norwegian lemming (Lemmus lemmus) at Kilpisjärvi (NW Finnish Lapland) during a moderate population peak. Annal Zool Fenn 39: 335–341.

Virtanen R, Eskelinen A & Gaare E (2003) Long-term changes in alpine plant communities in Norway and Finland. In: Nagy L, Grabherr GC, Körner C & Thompson DBA (eds) Alpine Biodiversity in Europe. Berlin, Springer: 411–422.

Wallenstein MD, McMahon SK & Schimel JP (2009) Seasonal variation in enzyme activities and temperature sensitivities in Arctic tundra soils. Glob Chang Biol 15: 1631–1639.

Wal R van der (2006) Vertebrate herbivores and ecosystem control: cascading effects of faeces on tundra ecosystem. Ecogr 27: 242–252.

Wal R van der & Brooker RD (2004) Bryophytes mediate grazer impacts on grass abundance in arctic ecosystems. Funct Ecol 18: 77–86.

Wardle DA, Bardgett RD, Klironomos JN, van der Putten WH & Wall DH (2004) Ecological linkages between aboveground and belowground biota. Science 304: 1629–1633.

Williams M, Street LE, van Wijk MT & Shaver GR (2006) Identifying differences in carbon exchange among arctic ecosystem types. Ecosystems 9: 288–304.

Zimov S, Chuprynin I, Oreshko AP, Chapin FSI, Reynolds JF & Chapin MC (1995) Steppe-tundra transition: a herbivore-driven biome shift at the end of the Pleistocene. Am Nat 146: 765–794.

51

Original publications

This thesis is based on the following publications, which are referred throughout

the text by their Roman numerals:

I Ylänne H, Olofsson J, Oksanen L & Stark S (Manuscript) Grazer-induced grassification of deciduous shrub tundra may promote soil carbon sequestration.

II Ylänne H, Kaarlejärvi E, Väisänen M, Männistö M, Ahonen S, Olofsson J & Stark S (Manuscript) Fertilization mitigates warming-induced carbon loss in long-term grazing regimes but not in short-term exclosures.

III Ylänne H, Stark S, & Tolvanen A (2015) Vegetation shift from deciduous to evergreen dwarf shrubs in response to selective herbivory offsets carbon losses: evidence from 19 years of warming and simulated herbivory in the sub-arctic tundra. Glob Chang Biol 21: 3696–3711.

Paper III is reprinted with permission from John Wiley and Sons.

Original publications are not included in the electronic version of the dissertation.

52

A C T A U N I V E R S I T A T I S O U L U E N S I S

Book orders:Granum: Virtual book storehttp://granum.uta.fi/granum/

S E R I E S A S C I E N T I A E R E R U M N A T U R A L I U M

673. Pentinsaari, Mikko (2016) Utility of DNA barcodes in identification anddelimitation of beetle species, with insights into COI protein structure across theanimal kingdom

674. Lassila, Toni (2016) In vitro methods in the study of reactive drug metaboliteswith liquid chromatography / mass spectrometry

675. Koskimäki, Janne (2016) The interaction between the intracellular endophyticbacterium, Methylobacterium extorquens DSM13060, and Scots pine (Pinus sylvestrisL.)

676. Ronkainen, Katri (2016) Polyandry, multiple mating and sexual conflict in a waterstrider, Aquarius paludum

677. Pulkkinen, Elina (2016) Chemical modification of single-walled carbon nanotubesvia alkali metal reduction

678. Runtti, Hanna (2016) Utilisation of industrial by-products in water treatment :carbon-and silicate-based materials as adsorbents for metals and sulphate removal

679. Suoranta, Terhi (2016) Advanced analytical methods for platinum group elements :applications in the research of catalyst materials, recycling and environmental issues

680. Pesonen, Janne (2016) Physicochemical studies regarding the utilization of wood-and peat-based fly ash

681. Kelanti, Markus (2016) Stakeholder analysis in software-intensive systemsdevelopment

682. Ahmad, Muhammad Ovais (2016) Exploring Kanban in software engineering

683. Mustonen, Kaisa-Riikka (2016) Climate change and boreal rivers : predictingpresent-day patterns and future changes in hydrological regime and its effects onriver communities

684. Karppinen, Pasi (2016) Studying user experience of health behavior changesupport systems : a qualitative approach to individuals’ perceptions of web-basedinterventions

685. Sarja, Jari (2016) Developing technology pushed breakthroughs : defining andassessing success factors in ICT industry

686. Taušan, Nebojša (2016) Choreography modeling in embedded systems domain

UNIVERSITY OF OULU P .O. Box 8000 F I -90014 UNIVERSITY OF OULU FINLAND

A C T A U N I V E R S I T A T I S O U L U E N S I S

University Lecturer Tuomo Glumoff

University Lecturer Santeri Palviainen

Postdoctoral research fellow Sanna Taskila

Professor Olli Vuolteenaho

University Lecturer Veli-Matti Ulvinen

Planning Director Pertti Tikkanen

Professor Jari Juga

University Lecturer Anu Soikkeli

Professor Olli Vuolteenaho

Publications Editor Kirsti Nurkkala

ISBN 978-952-62-1509-9 (Paperback)ISBN 978-952-62-1510-5 (PDF)ISSN 0355-3191 (Print)ISSN 1796-220X (Online)

U N I V E R S I TAT I S O U L U E N S I SACTAA

SCIENTIAE RERUM NATURALIUM

U N I V E R S I TAT I S O U L U E N S I SACTAA

SCIENTIAE RERUM NATURALIUM

OULU 2017

A 687

Henni Ylänne

HERBIVORY CONTROL OVER TUNDRA CARBON STORAGE UNDER CLIMATE CHANGE

UNIVERSITY OF OULU GRADUATE SCHOOL;UNIVERSITY OF OULU,FACULTY OF SCIENCE;UNIVERSITY OF LAPLAND,ARCTIC CENTER

A 687

ACTA

Henni Ylänne