Climate change and biotic interactions in plant communities: effects on plant recruitment and growth, population dynamics and community properties

Klimaendringer og samspillet mellom arter i plantesamfunn: effekter på rekruttering og plantevekst, populasjonsdynamikk og samfunnsegenskaper

Philosophiae Doctor (PhD) Thesis Siri Lie Olsen Department of Ecology and Natural Resource Management Faculty of Environmental Science and Technology Norwegian University of Life Sciences Ås 2014

Thesis number 2014:83 ISSN 1894-6402 ISBN 978-82-575-1242-2

PhD supervisors Associate professor Kari Klanderud Department of Ecology and Natural Resource Management, Norwegian University of Life Sciences, Norway Professor Vigdis Vandvik Department of Biology, University of Bergen, Norway Senior Researcher Olav Skarpaas Norwegian Institute for Nature Research, Norway

Evaluation committee Professor Greg H. R. Henry Department of Geography, University of British Columbia, Canada Professor Bente J. Graae Department of Biology, Norwegian University of Science and Technology, Norway Professor Tone Birkemoe Department of Ecology and Natural Resource Management, Norwegian University of Life Sciences, Norway

ii

“We have reason to believe that species in a state of nature are limited in their ranges by the competition of other organic beings quite as much as, or more than, by adaptation to particular climates” (Charles Darwin in On the origin of species by means of natural selection, 1859)

iii

iv

Acknowledgements As a bachelor and master student doing fieldwork at Finse I daydreamed about being allowed to work with the open-top chambers. Actually doing it one day was more than I had hoped for, and I am extremely proud to be part of ITEX. Thank you for giving me the opportunity, Kari. You know that you were my idol as a master student? I used your papers as guidance when writing my master thesis, hoping to some day master the art of scientific writing. You have also been the best supervisor a PhD student can ask for. When the other PhDs complained about their supervisors, I could never join in, because I had nothing to complain about! You had a plan for my PhD work from the very beginning, I could knock on your door whenever I had a question, and I always got comments back on my work if I asked for it. You serve as an example for other supervisors. Thank you for always having time, for always believing in me, and for everything you have taught me. I am also very proud to be a member of the SeedClim team. Great experiment, great people. I am especially thankful for my two co-supervisors: Thank you,Vigdis, for always being enthusiastic and full of energy and new ideas – while at the same time being a brilliant scientist. SeedClim is in the best hands with you. Your enthusiasm is contagious! Olav, thank you for bringing me back down to Earth with your excellent analytical skills and critical thinking. Thanks also for introducing me to plant demography, and to NINA. Joachim, Pascale, Eric, Christine, John, Serge… This journey would not have been the same without you. Thank you for all the long, warm, cold, wet, fun and exhausting days in the field. Lise, thanks for involving me in the tree seedling recruitment study. Thanks also to all our hardworking field assistants: Ali, Annick, Bernarda, Fanney, Ingeborg, Izzie, Jeanette, Monica, Siobhan – as well as Aud, Amund, Dani, Sigrid, Mari and Owen for helping out with the graminoid removal. And Richard, I would still be struggling to extract my data from the SeedClim database if it wasn’t for you. Thanks also to all my colleagues at INA, especially those of you I have had the pleasure of teaching with: John, Christian, Massimo, Mariken, Camilla, Markus… I am very grateful to Kari and Yngvar for letting me teach; it has been a great joy to me. And thanks to all the INA students for giving me a welcome break from PhD worries and letting me introduce you to the wonderful world of plants. I don’t know if you see it yet, but botany is more fun than zoology! Thanks also to Johan and Rebekka for lunches and chats. v

Finally, I would like to thank my family and my closest friends: mum and dad, Turid, and Therese, Marit, Cathrine, Dani and Anette. Your support has been essential for the completion of this thesis. I also wish to thank my dog Linnea for the good company and for making me go for long walks every day to clear my head. I am very grateful to everyone who has looked after her while I have been away on fieldwork: Ole Kristian, Anna Marie and Bjørn, Lars Erik and Stine… The summer logistics would have been impossible without you. Thanks also to Olga, who went out of her way to take care of my garden. My dearest Torbjørn. I don’t know what to say… Thank you for moving to Ås with me, even though you didn’t really want to, so that I could start my PhD. You have supported me from day one, and I could not have done this without you. We started this together, and I wish we could have finished it together.

vi

Table of contents

List of papers

ix

Summary (in English)

xi

Sammendrag (på norsk)

xiii

Synopsis

1

Introduction

1

Materials and methods summary

4

SeedClim climate grid

4

Finse ITEX site

8

Main results

11

Discussion

18

Climate alters biotic interactions

18

A common trend

19

Consequences of increased competition

20

Concluding remarks

21

Acknowledgements

21

References

22

Paper I Paper II Paper III Paper IV Paper V

vii

viii

List of papers This thesis consists of the following papers (hereafter referred to by their roman numeral): Paper I Tingstad, L., Olsen, S. L., Klanderud, K., Vandvik, V. and Ohlson, M. The role of temperature, precipitation and competition as determinants of tree seedling recruitment across the tree line ecotone. Submitted to Oecologia. Paper II Olsen, S. L., Töpper, J. P., Skarpaas, O., Vandvik, V. and Klanderud, K. From facilitation to competition: climate warming shifts dominant plant interactions in semi-natural grasslands. Submitted to Ecology. Paper III Olsen, S. L., Vandvik, V., Guittar, J. and Klanderud, K. Climate-driven changes in biotic interactions affect seedling recruitment and species richness in semi-natural grasslands. Manuscript. Paper IV Olsen, S. L. and Klanderud, K. 2014. Biotic interactions limit species richness in an alpine plant community, especially under experimental warming. Oikos 123: 71-78. Paper V Olsen, S. L. and Klanderud, K. 2014. Exclusion of herbivores slows down recovery after experimental warming and nutrient addition in an alpine plant community. Journal of Ecology 102: 1129-1137. Published papers are re-printed with the permission of the publishers.

ix

x

Summary Global climate change is already affecting the world’s biota. In addition to the direct effects of increased temperatures and altered precipitation patterns, climate change can affect plants and animals by altering biotic interactions. In this thesis, which consists of studies from seminatural grasslands and alpine Dryas heath in southern Norway, I examine how changes in temperature and precipitation affect interactions between plants at different levels of organization in plant communities: individual plants, populations and the community as a whole. I also examine how interactions between trophic levels influence plant community recovery from climate warming. Using a combination of experimental warming and studies along natural climate gradients I show that changes in climate, including both temperature and precipitation, can alter biotic interactions in plant communities. Moreover, there was a common trend across organizational levels: increasing competition with increasing temperature. Increased competition between plants in a warmer future climate may both enhance and impede the direct effects of temperature increase. On one hand, increased competition may slow down tree-line advancement and inhibit the establishment of invasive species. On the other hand, increased competition may prevent migrating species from tracking their climatic niche, and reduce recruitment, population growth rates and ultimately species richness of the resident vegetation. Whereas there was an overall response to temperature increase across organizational levels, the effect of precipitation was more complex, indicating that simultaneous changes in precipitation patterns may modify the response to climate warming. Moreover, the effects of increased competition may be moderated by interactions between trophic levels. Herbivory may for instance prevent the expected increase in competition and increase reversibility of warming-induced changes in plant communities. In semi-natural ecosystems such as grasslands, traditional grazing or mowing activities may play a similar role. In conclusion, climate warming, and to a lesser degree altered precipitation patterns, will affect species interactions at all levels of organization in plant communities, which may modify the direct effect of climate change. The identification of a general effect of climate warming on biotic interactions in plant communities can improve predictions of climate change impacts on ecosystems.

xi

xii

Sammendrag Globale klimaendringer påvirker allerede livet på jorda. I tillegg til direkte effekter av økt temperatur og endrede nedbørmønstre, påvirker klimaendringene planter og dyr ved å forskyve samspillet mellom dem. I denne avhandlingen, som består av studier fra ugjødslet beitemark og alpin reinrosehei i Sør-Norge, ser jeg på hvordan endringer i temperatur og nedbør påvirker samspillet mellom planter på de ulike organisasjonsnivåene i et plantesamfunn: individer, populasjoner og samfunnet som helhet. I tillegg ser jeg på hvordan samspillet mellom ulike trofiske nivåer kan påvirke plantesamfunnets evne til å returnere til det opprinnelige etter en temperaturøkning. Ved hjelp av en kombinasjon av eksperimentell oppvarming og studier langs naturlige klimagradienter viser jeg at endringer i klima, både temperatur og nedbør, kan påvirke samspillet i et plantesamfunn. Jeg identifiserer også en felles trend på tvers av organisasjonsnivåer: økende konkurranse med økende temperatur. Økt konkurranse mellom planter i et varmere klima kan både fremme og hemme direkte effekter av oppvarming. På den ene siden kan økt konkurranse motvirke heving av tregrensa og redusere etablering av invaderende arter. På den annen side kan økt konkurranse forhindre arter på vandring fra å holde tritt med sine klimatiske nisjer, samt redusere rekruttering, populasjonsvekstrater og dermed artsrikdom i stedegen vegetasjon. Mens jeg fant en generell effekt av økt temperatur på tvers av organisasjonsnivåer, sprikte virkningen av nedbørvariasjon i mye større grad, noe som tyder på at endringer i nedbørmønstre kan modifisere effekten av oppvarming. Videre kan effekten av økt konkurranse motvirkes av samspill mellom ulike trofiske nivåer. For eksempel kan beite forhindre den forventede økningen i konkurranse, samt reversere allerede oppståtte effekter av oppvarming. I menneskepåvirkede økosystemer som enger kan tradisjonell hevd i form av beite eller slått ha en tilsvarende virkning. Kort oppsummert vil et varmere klima, og i mindre grad endringer i nedbørmønstre, påvirke samspillet mellom arter på alle de ulike nivåene i et plantesamfunn og derved modifisere direkte effekter av klimaendringer. Funnet av en generell effekt av oppvarming på biotiske interaksjoner i plantesamfunn kan forbedre vår evne til å forutsi klimaendringenes innvirkning på økosystemer.

xiii

xiv

Synopsis

xv

xvi

Introduction Climate change is inevitable. As extensively documented by the Intergovernmental Panel for Climate Change, anthropogenic emissions of greenhouse gases are resulting in global climate change, which includes both global temperature increase and altered precipitation patterns (IPCC 2013). This alteration of the world’s climate has been predicted to have large consequences for ecosystems worldwide, and climate change is considered one of the major threats to global biodiversity (e.g. Sala et al. 2000, Thuiller et al. 2007, Bellard et al. 2012). Thus, “predicting and ameliorating the consequences of climate change presents a major challenge for ecologists” (Gilman et al. 2010). Decades of research has generated vast knowledge about the effects of climate change on the world’s biota (reviewed in e.g. Walther et al. 2002, Parmesan 2006). Climate warming has for instance been linked to altered phenology (e.g. Root et al. 2003, Parmesan and Yohe 2003, Visser and Both 2005, Cleland et al. 2007), northward and upward migrations (e.g. Grabherr et al. 1994, Lenoir et al. 2008, Parmesan and Yohe 2003, Pauli et al. 2012, Harsch et al. 2009) and shifts in local species abundances (e.g. Sturm et al. 2001, Thomas et al. 2006, Kausrud et al. 2008, Elmendorf et al. 2012b) of both plants and animals. Such changes may in turn affect ecosystem structure and function and thereby ultimately ecosystem services (e.g. Chapin et al. 1997, Mooney et al. 2009, Montoya and Raffaelli 2010). Climate change may influence plant and animals directly through physiological responses to temperature and precipitation change, but also indirectly through altered biotic interactions (e.g. Adler et al. 2012), which may modify the direct effects of climate. Climate change has been predicted to alter all major types of biotic interactions (Tylianakis et al. 2008), and climate-driven shifts in biotic interactions within and among trophic levels have been found in mammals, birds, fish, amphibians, reptiles, invertebrates (zooplankton, insects, spiders), plants, fungi and bacteria (Gellesch et al. 2013 and references therein, Post 2013 and references therein). However, many authors point out that we still lack understanding of the indirect effects of climate. More than twenty years ago, Kareiva et al. (1993) wrote that “Missing, however, has been an examination of how species interactions (…) might exacerbate or mitigate the dire environmental consequences expected with global warming”, and this has been repeated many times since (e.g. Brooker 2006, Gilman et al. 2010, Lavergne et al. 2010, van der Putten et al. 2010, Adler et al. 2012, Urban et al. 2012, Post 2013). 1

Plants form the basis of terrestrial food webs, and climate-driven changes to plant communities may therefore have cascading effects onto other trophic levels. Thus, understanding how plants respond to climate variation is of major importance for predicting how terrestrial ecosystems are affected by global climate change. Although several recent studies have examined how changes in climate alter plant-plant (e.g. Callaway et al. 2002, Maestre et al. 2005, He et al. 2013, Anthelme et al. 2014, Michalet et al. 2014, Soliveres and Maestre 2014) and plant-animal interations (e.g. Memmott et al. 2007, Post and Pedersen 2008, Hegland et al. 2009, Gillespie et al. 2013, Post 2013), we still lack knowledge of how climate-driven changes in biotic interactions influence the response of plants and plant communities to climate warming and precipitation changes. This thesis is a contribution to filling that knowledge gap. Most studies of climate change effects focus solely on warming, but it has been difficult to detect a uniform response of plant communities to climate warming at large spatial scales (e.g. Elmendorf et al. 2012ab, Grytnes et al. 2014). This lack of large-scale patterns may be attributed to the complex interplay between temperature and precipitation in climate change impacts on plants (e.g. Luo et al. 2008, Elmendorf 2012ab, Harsch and HilleRisLambers 2014). Thus, it is crucial to examine how biotic interactions are affected by simultaneous changes in temperature and precipitation. Moreover, most climate change studies are conducted using either experimental climate manipulations or studies along environmental gradients. While experiments are important for inferring underlying mechanisms behind plant responses to climate change, they are often restricted to short time-spans and small spatial scales (Dunne et al. 2003, 2004), and the experimental design may have unwanted side-effects (e.g. De Frenne 2014). Gradient studies can be conducted at large spatial scales and reflect long-term adaptations to climate, but underlying mechanisms are more difficult to identify (Dunne et al. 2003, 2004). Combining experiments with gradient studies makes it possible to overcome the limitations of both methods and thereby produce more robust conclusions (Rustad et al. 2001, Dunne et al. 2003). Until now, most studies examining biotic interactions in plant communities have focused on individual plants (e.g. Choler et al. 2001) or a few selected species (e.g. Klanderud 2005). Although community-level studies are scarce, several authors warn against scaling up species2

level results to predict effects on higher levels (Pace 1993, Tylianakis et al. 2008, Anthelme et al. 2014, Soliveres and Maestre 2014, Soliveres et al. 2014). Despite this warning, the effect of climate variation on biotic interactions between plants seems surprisingly consistent between different organizational levels, at least within climatic regions (e.g. Callaway et al. 2002, Michalet et al. 2014, Soliveres and Maestre 2014). Yet, I don’t know of any attempts to examine whether climate-driven alterations in biotic interactions affect different levels of organization in contrasting ways, or whether it is possible to find a general pattern. The overall aim of this thesis is to examine the effect of climatic change on biotic interactions in plant communities and try to identify common trends among different levels of organization. Examining multiple organizational levels and plant life stages using a combination of experimental and gradient approaches, I present an extensive study of how climate, including both temperature and precipitation change, affects biotic interactions in plant communities within a climatic region. Finally, I examine how interactions between trophic levels may influence the reversibility of warming-induced vegetation changes. The specific questions addressed in this thesis are: -

How does climate influence the effect of biotic interactions on recruitment from seed and growth of individual plant seedlings (paper I)?

-

How does climate influence the effect of biotic interactions on plant population dynamics (paper II)?

-

How does climate influence the effect of biotic interactions on plant community properties such as species richness, composition and functional traits (paper III, IV)?

-

How do biotic interactions between trophic levels affect the reversibility of warminginduced changes in plant communities (paper V)?

3

Materials and methods summary I used two different study systems to examine the effect of climate change on biotic interactions in plant communities: the SeedClim climate grid (paper I, II and III) and the Finse ITEX site (paper IV and V) in southern Norway. SeedClim climate grid The SeedClim climate grid was established as part of the SeedClim climate change project and consists of twelve semi-natural grassland sites situated in the fjord landscape of southern Norway (Fig. 1A-B). The sites were selected to fit within a climate grid composed of a natural temperature gradient (from alpine to lowland) replicated along a precipitation gradient (from continental to oceanic). This unique climate grid combines three levels of summer temperature (mean of the four warmest months; alpine 6.5°C, intermediate 8.5°C and lowland ca. 10.5°C) with four levels of annual precipitation (ca. 600, 1200, 2000 and 2700 mm) (Fig. 2). All study sites are grazed, species-rich grasslands on south-facing, shallow slopes with calcareous bedrock. The sites are fenced to avoid animal disturbance, but are mowed annually. Details on site selection and characteristics can be found in Meineri et al. (2013, 2014). Many different studies are carried out in the SeedClim study sites, all with the aim of examining the effects of global climate change on plant communities and ecosystems using natural climate gradients in combination with experiments (e.g. Bargmann 2009, Berge 2010, Pötsch 2010, Fariñas 2011, Boixaderas 2012, Tingstad 2012, Meineri et al. 2013, 2014, Skarpaas et al. submitted, Klanderud et al. submitted, Michel et al. in prep., Töpper et al. in prep., Vandvik et al. in prep.).

4

B

A

SeedClim

Fig. 1. Map showing the location of the SeedClim climate grid in southern Norway (A) and an example of species-rich semi-natural grassland vegetation in the warmest and driest study site (B). Photo: Siri Lie Olsen.

Fig. 2. A schematic overview of the SeedClim climate grid in southern Norway with the study sites (grey squares) positioned along temperature and precipitation axes.

5

In paper I we examined how climate-driven variation in biotic interactions affected seedling emergence, establishment and growth of individual plants. Seeds of the boreal tree species Pinus sylvestris and Picea abies were sown in intact vegetation and on bare soil in experimentally created gaps in each of the study sites. All seedlings were counted the following spring to determine seedling emergence and again in the fall to determine seedling establishment. The seedlings were then harvested to examine the effects of climate and biotic interactions on growth and allocation patterns. Mixed effects models were used to examine how climate and biotic interactions affected seedling emergence, establishment, height and biomass. In paper II and III we examined how climate-driven variation in interactions between graminoids and forbs affected population dynamics of four study species and community properties, respectively. In both studies graminoids were experimentally removed (Fig. 3A) to determine how a dominant functional group affects co-occurring plant species. In paper II detailed demographic studies of four focal species (Fig. 3B-E) were conducted and analyzed using integral projection models (IPMs). Population growth rates in removal and control plots within each site were compared using Wilcoxon rank sum tests, and a life table response experiment (LTRE) was used to determine which vital rates (survival, growth, clonality, fecundity) contributed most to the differences in population growth rates between removal and control plots. In paper III we examined how experimental graminoid removal affected seedling recruitment and plant community properties (species richness, evenness and composition and functional traits) of subordinate species, and how this effect varied with temperature and precipitation. Following vegetation analyses, seedling recordings and trait collection, mixed effects models were used to examine how climate and biotic interactions affected non-graminoid plant species richness and evenness, functional traits (plant height, leaf size, seed mass and specific leaf area) and the number of seedlings. The effects on nongraminoid species composition were determined using ordination.

6

A

B

C

D

E

Fig. 3. An example of an experimental plot where graminoids have been removed (A) and the four focal species of paper II, Viola palustris (B), Viola biflora (C), Veronica officinalis (D) and Veronica alpina (E), in semi-natural grasslands in southern Norway. Photos: Siri Lie Olsen.

7

Finse ITEX site Finse is situated in the alpine region of southern Norway (Fig. 4A) and has a slightly oceanic climate (Moen 1998) with a mean monthly temperature and rainfall during summer (JuneAugust) of 6.3 ºC and 89 mm, respectively (Norwegian Meteorological Institute 2010). The study site is located on the southwest-facing slope of Mt. Sanddalsnuten (peak at 1556 m a.s.l.) where the bedrock consists mainly of phyllite, supporting a species rich Dryas heath community dominated by Dryas octopetala (Fig. 4B). The study site is further described in Klanderud (2005, 2008, 2010) and Klanderud and Totland (2005ab, 2007, 2008). The Finse study site is included in the International Tundra Experiment (ITEX), a global network of experimental sites where researchers examine the effects of global climate change on high altitude and latitude ecosystems. The common feature of the ITEX sites is the use of open-top chambers (OTCs) to experimentally manipulate temperature (see e.g. Marion et al. 1997). At Finse OTCs with an inside diagonal of 1 m, which increase summer air temperature by approximately 1.5 °C (Klanderud 2005, Klanderud and Totland 2005b), have been used for experimental warming of vegetation since 2000 (Fig. 4B).

A

B

Finse

Fig. 4. Map showing the location of Finse in southern Norway (A) and an example of the Dryas heath vegetation with open-top chambers (B). Photo: Siri Lie Olsen.

8

In paper IV we examined the effects of experimental warming on long-term species establishment following a seed-addition experiment established by Klanderud and Totland (2007). Moreover, we examined the influence of different biotic and abiotic factors on species richness in the long term. Klanderud and Totland (2007) added seeds of 27 species to the vegetation inside the OTCs (Fig. 5A) and in control plots, and based on vegetation analyses (Fig. 5B) species richness was assessed at the start of the experiment and after four and eleven years. In addition, Klanderud and Totland (2007) recorded biotic and abiotic variables in the experimental plots. The effect of warming and seed addition over time was examined using mixed effects modelling, and the relationship between the change in species richness and the different biotic and abiotic variables in the short and long term was assessed by multiple regressions.

A

B

Fig. 5. Detailed photo of an open-top chamber used for experimental warming (A) and an ambient temperature plot with the frame used for vegetation analyses (B) at Finse, Norway. Photos: Siri Lie Olsen. In paper V we examined the recovery of the Dryas heath community, as well as the effect of herbivory on the recovery process, following a global change experiment established by Klanderud and Totland (2005b). After seven years of experimental warming and nutrient addition, treatments were ceased and herbivore exclosures (Fig. 6) were erected upon half of the study plots. Vegetation analyses were carried out prior to global change treatments and after three years of treatments (Klanderud and Totland 2005b), shortly after treatment cessation and six years after treatment cessation, allowing for studies of plant community 9

recovery from global change. The recovery of plant community composition was examined using various ordination techniques, whereas mixed effects models were used to assess recovery of abundance and species richness of the main functional groups, as well as litter cover.

Fig. 6. Herbivore exclosures made from galvanized net with a mesh size of 12.7 × 12.7 mm at Finse, Norway. Photo: Siri Lie Olsen.

10

Main results Paper I showed that competition from the resident vegetation limited plant recruitment and growth in the semi-natural grasslands sites, as tree seedling emergence (Fig. 7A) and establishment (Fig. 7B) were higher in gaps than in intact vegetation. Moreover, individual seedlings were shorter, but had a higher biomass in gaps (not shown). The positive effect of gaps on seedling emergence, establishment and biomass increased with increasing temperature, indicating increasing competition intensity with climate warming. Precipitation also influenced the effect of gaps, which was most pronounced at intermediate precipitation levels (Fig.7A-B).

A

B

A

B

Fig. 7. Emergence (A) and establishment (B) of pine seedlings in intact vegetation (black bars) and experimentally created gaps (grey bars) at different levels of temperature (warm to cold: lowland, intermediate, alpine) and precipitation (dry to wet: 1-4) in semi-natural grasslands in southern Norway. Adapted from paper I, where figures for spruce can be found.

11

Paper II showed that biotic interactions with graminoids affected population growth rates of co-occurring forbs in semi-natural grasslands, but that the net outcome of these interactions varied with climate. For two of the four study species the net interaction between graminoids and forbs switched from facilitation at low temperatures to competition at high temperatures (Fig. 8), while the negative effect of competition on population growth rates increased with temperature for a third species (not shown). The increase in competition with temperature was mainly due to negative effects of graminoids on forb survival at high temperatures. Precipitation influenced the strength of interactions, but the effect was temperature- and species-specific (Fig. 8).

A

B

Fig. 8. The difference in population growth rates (∆λ) between removal and control plots for Viola palustris (A) and Viola biflora (B) at different levels of temperature (cold to warm: alpine, intermediate, lowland) and precipitation (four levels from dry to wet) in semi-natural grasslands in southern Norway. Values above the zero line indicate competition, while values below the zero line indicate facilitation. Error bars show 95 % confidence intervals. Adapted from paper II.

12

Paper III showed that competition from dominant graminoids limited seedling recruitment (not shown) and richness (Fig. 9) of subordinate plant species in the majority of the seminatural grassland sites. However, interactions with graminoids also varied with climate: Competition seemed to increase with temperature, as the net outcome of plant interactions varied from competition to facilitation along the precipitation gradient in the cold alpine, while competition dominated at all precipitation levels in the warm lowlands (Fig. 9). There were no effects of graminoid removal on non-graminoid species composition or functional traits (not shown).

Fig. 9. The difference in non-graminoid plant species richness between the graminoid removal and control plots at different levels of temperature (cold to warm: alpine, intermediate, lowland) and precipitation (four levels from dry to wet) in semi-natural grasslands in southern Norway in 2013. Values above the zero line indicate competition, while values below the zero line indicate facilitation. To account for initial differences in species richness, 2011 species richness was subtracted from the 2013 data. Data from 2012 are not shown. Error bars show means ±1 SE. Adapted from paper III.

13

Paper IV showed that seed addition increased short-term species recruitment in the alpine Dryas heath, but this initial increase was cancelled out by a corresponding decrease in species richness in the long term (Fig. 10). Whereas the short-term increase in species richness was related to both biotic and abiotic factors, the subsequent decrease was related to biotic factors only (Table 1), indicating increased influence of biotic interactions over time. The relative importance of biotic interactions also seemed to increase with experimental warming, as there were more significant relationships with biotic factors in the warmed plots (Table 1).

Fig. 10. Vascular plant species richness in 2000, 2004 and 2011 with (triangles) and without (circles) seed addition under ambient temperature (open symbols) and experimental warming (filled symbols) at Finse, Norway. Error bars show mean values ±1 SE. Adapted from paper IV.

14

Tabel 1. Regression coefficients for multiple regressions between the change in vascular plant species richness between 2000 and 2004 (Klanderud and Totland 200) and 2004 and 2011 (paper IV) and different environmental variables in experimentally warmed and ambient temperature plots at Finse, Norway. Some variables were not included in the regressions (-) by Klanderud and Totland (2007). In the current study non-significant (ns) variables were excluded during model selection. *p < 0.05, **p < 0.01, ***p < 0.001. Adapted from paper IV. 2000-2004 Environmental variables No. of species added Shannon’s diversity index

Warming

2004-2011

Ambient

0.51***

0.53***

-

-

Warming -0.27***

Ambient -0.30***

-12.38***

ns

Graminoid species richness

-0.54***

-0.39**

8.96**

ns

Woody species richness

0.05

-0.01

-1.36*

ns

ns

-0.61*

Forb species richness

-0.31

-0.31**

Litter cover

-

-

-0.98*

ns

Total vegetation cover

-

-

ns

ns

Dryas cover

-0.29**

-0.40**

ns

ns

Extractable soil N

-0.37**

-0.42**

ns

ns

Soil moisture

0.61***

0.40**

ns

ns

Bare soil cover

0.20

0.16

ns

ns

15

Paper V showed that recovery of the Dryas heath community from experimental warming and nutrient addition is a slow process. A persistent shift in species composition, especially following nutrient addition and warming combined with nutrient addition, seemed to inhibit recovery and maintain a community dominated by highly competitive grasses at the expense of the previously dominating dwarf shrub Dryas octopetala, lichens and bryophytes (Fig. 11 and 12). The effects of warming were weaker and more transient than those including nutrient addition. Excluding herbivores decreased the community recovery rate, seemingly by maintaining unfavourable conditions for lichens and bryophytes (Fig. 11).

Fig. 11. Principal response curve (PRC) ordination of mean plant community composition in control, warming, nutrient addition and warming combined with nutrient addition plots during the treatment application (2000-2007) and subsequent recovery (2007-2012) with herbivores present (dashed line) and herbivores excluded (solid line) in an alpine Dryas heath at Finse, Norway. Treatments were ceased and herbivore exclosures were erected in spring 2007 (vertical dotted line). The horizontal grey line represents control plots with herbivores present, to which all other treatments are compared. Note that only PRC axis 1 is shown. The responses of the most common vascular plants (V), lichens (L) and bryophytes (B) are shown to the right. Adapted from paper V. Full species names can be found in the supporting information to paper V. 16

A

B

C

D

Fig. 12. Six years after treatment cessation there was still a clearly visible difference in the plant community composition of the control plots (A) compared to the warming (B), nutrient addition (C) and warming combined with nutrient addition (D) plots at Finse, Norway. The superimposed net used for vegetation analyses has a mesh size of 5 × 5 cm.

17

Discussion Together paper I-V show that changes in climate, including both temperature and precipitation, can alter biotic interactions at all organizational levels of plant communities, from individual plants and populations to the entire community. While the effect of climate variation, especially precipitation, differed between organizational levels, there was also a common trend: a shift towards increased competition with increasing temperature. This warming-induced increase in competition may both enhance and impede the direct effect of rising temperatures. Climate alters biotic interactions As discussed by for instance Tylianakis et al. (2008), climate change is predicted to alter biotic interactions in natural ecosystems, thereby modifying the direct effects of increased temperature and altered precipitation patterns. Accordingly, both temperature and precipitation change influenced interactions between plants in the semi-natural grassland sites. Interactions between tree seedlings and their neighbours (paper I), between graminoids and four co-occurring forb species (paper II) and between graminoids and non-graminoid grassland species (paper III) all varied within the climate grid, although more strongly along the temperature gradient than the precipitation gradient. Similarly, experimental warming influenced plant-plant interactions in the alpine Dryas heath by increasing the relative importance of biotic interactions for long-term species establishment (paper IV). In addition, altered biotic interactions between vascular and non-vascular plants seemed to maintain warming-induced changes in community composition (paper V). My findings of climate-effects on interactions between plants are in line with other studies along climate gradients (e.g. Callaway et al. 2002, Maestre et al. 2005, Anthelme et al. 2014, Michalet et al. 2014, Soliveres and Maestre 2014) and using experimental warming (e.g. Klanderud 2005, Klanderud and Totland 2005a) or water manipulations (e.g. Suttle et al. 2007) showing that temperature and precipitation changes may alter plant-plant interactions. Results seem consistent, regardless of method and study system. Together, our studies show that altered biotic interactions can influence the response of plants to climate change and therefore need to be taken into consideration when assessing the impacts of climate change on plant communities and ecosystems (see e.g. Tylianakis et al. 2008, Gilman et al. 2010).

18

A common trend Although several authors warn against scaling up the effects of climate change from one organizational level to another (e.g. Soliveres and Maestre 2014), I found a common trend in the influence of climate on biotic interactions between individual plants, populations and communities: an increase in competition with increasing temperature. This enhancement of competition with climate warming was evident for emergence, early establishment and growth of individual seedlings (paper I), population growth rates (paper II) and community properties such as species richness (paper III, IV), and the trend was the same for both experimental warming (paper IV) and along a natural temperature gradient (paper I, II, III). Increased competition with increasing temperature has previously been found using both experimental warming (Dunnett and Grime 1999, Klanderud 2005) and studies along climate gradients (e.g. Callaway et al. 2002, Kikvidze et al. 2005, see also Anthelme et al. 2014). My findings support the stress-gradient hypothesis, which states that the balance between the two dominant plant-plant interactions, competition and facilitation, should vary along environmental gradients, with the relative importance of competition increasing as environmental stress decreases (i.e. with increasing productivity, sensu Grime 2001) (e.g. Brooker and Callaghan 1998). In relatively cold climates such as in my study systems, increased temperature most likely reduces environmental stress for plants, thereby increasing productivity and intensifying competition. Whereas there seemed to be an over-all trend in the effect of temperature on the outcome of biotic interactions, the results for precipitation were more complex. While competition intensity peaked at intermediate precipitation levels for emergence and establishment of individual tree seedlings (paper I), the effect of precipitation on changes in population growth rates seemed to be temperature- and species-specific (paper II). For non-graminoid seedling recruitment and species richness, precipitation had a greater influence under low than high temperature (paper III). Thus, I did not find a common trend in the response to precipitation. Studies from arid environments show that competition increase with increasing rainfall (e.g. Pugnaire and Luque 2001, Holzapfel et al. 2006, Armas et al. 2011; see also discussion in He et al. 2013) as water-stress is relieved and productivity increase. However, this relatively linear relationship is probably not the same in the more mesic climate of southern Norway, where drought is less common. Moreover, winter precipitation in the form of snow, including both absolute amounts and snow-cover duration, probably play a more

19

important role for plants at low than high temperatures (see paper III), thereby obscuring any across-temperature trends. Consequences of increased competition Increased competition in plant communities in a warmer future climate may both enhance and impede the direct effects of temperature increase. For instance, increased competition within the resident vegetation may slow down tree line advancements (paper I) and inhibit the establishment of invasive species (paper IV). On the other hand, reduced recruitment of new species into existing plant communities due to increased competition intensity may prevent migrating species from tracking their climatic niche (Urban et al. 2012), thereby making them more susceptible to the direct effects of a changing climate. Moreover, increased competition may favour highly competitive dominant species over subordinate species, thereby reducing seedling recruitment (paper I, III), population growth rates (paper II) and ultimately community species richness (paper III, IV) of the resident vegetation. Lichens and bryophytes, as well as low-stature forbs, seem especially sensitive to increased competition from dominants (paper V) (see also e.g. Cornelissen et al. 2001, Klanderud and Totland 2005b, Walker et al. 2006, Lang et al. 2012). Finally, shifts in interactions between plants may stabilize the changes induce by climate warming (paper V), thereby maintaining an altered community composition. Although short- and long-term studies showed a similar trend of increasing competition with climate warming, there was an apparent contradiction in the consequence of reduced seedling recruitment: Whereas increased competition may reduce seedling recruitment (paper I, III) and in turn limit species richness of the resident vegetation in the short term (paper III), recruitment from seed did not influence species richness in the long term (paper IV) and played a negligible role for population dynamics (paper II). However, only occasional recruitment events are needed to maintain populations of long-lived, clonally reproducing alpine and grassland species (e.g. Watkinson and Powell 1993). If increased competition reduces survival of adult plants (paper II) and further limits seedling recruitment (paper I, III), this may have negative consequences for population growth rates and thus species richness and possibly community structure and function in the long term (see also discussion of time lags in paper III).

20

Altered precipitation patterns may to some degree modify the magnitude of warming-induced increases in competitive (paper I, II, III). However, the interacting effect of temperature and precipitation seems highly complex and difficult to predict. The effect of increased competition may also be modified by interactions between different trophic levels. Herbivory has been shown to prevent climate-driven changes in plant communities by reducing dominance and thereby the expected increase in competition (e.g. Olofsson et al. 2009, Post and Pedersen 2008, Ravolainen et al. 2014). Moreover, herbivory may increase the reversibility of warming-induced changes in plant communities (paper V), and large herbivores can create small-scale disturbance which may promote seedling recruitment (see paper I, II). In semi-natural ecosystems such as grasslands, management interventions in the form of livestock grazing or mowing may play the same role as natural herbivores and reduce the impact of climate-driven increases in competition (see paper II, III). Concluding remarks This thesis shows that climate warming, and to a lesser degree altered precipitation patterns, will affect species interactions at all levels of organization in plant communities. Whereas I found an overall increase in competition with climate warming across organizational levels, the effect of precipitation was complex, which may obscure any uniform response to temperature increase at large spatial scales (e.g. Elmendorf et al. 2012ab, Grytnes et al. 2014). Further research is needed to understand how altered precipitation patterns, including winter precipitation in the form of snow, will affect biotic interactions in plant communities. Moreover, climate-driven shifts in interactions between trophic levels, for instance between plants and soil biota (e.g. Hågvar and Klanderud 2009), are not fully understood and deserve further attention. In conclusion, I show that it is possible to find a general effect of climate warming on biotic interactions in plant communities within a climatic region. Identifying such trends, and incorporating them into regional-scale assessments of climate change effects, can improve predictions of climate change impacts on ecosystems. Acknowledgements I wish to thank K. Klanderud, O. Skarpaas, V. Vandvik and Pascale Michel for helpful comments on earlier versions of this synopsis.

21

References Adler, P. B., H. J. Dalgleish, and S. P. Ellner. 2012. Forecasting plant community impacts of climate variability and change: when do competetive interactions matter? Journal of Ecology 100:478-487. Anthelme, F., L. A. Cavieres, and O. Dangles. 2014. Facilitation among plants in alpine environments in the face of climate change. Frontiers in Plant Science. DOI: 10.3389/fpls.2014.00387. Armas, C., S. Rodríguez-Echeverría, and F. I. Pugnaire. 2011. A field test of the stressgradient hypothesis along an aridity gradient. Journal of Vegetation Science 22:818827. Bargmann, T. 2009. How are plants responding to a changing climate? A case study of growth and allocation in Veronica alpina, Viola biflora, Veronia officinalis and Viola palustris in western Norway. Master thesis, University of Oxford, Oxford. Bellard, C., C. Bertelsmeier, P. Leadley, W. Thuiller, and F. Courchamp. 2012. Impacts of climate change on the future of biodiversity. Ecology Letters 15:365-377. Berge, A. 2010. Seedbank, seedrain and seedling recruitment along climate gradients in southern Norway. Master thesis, University of Bergen, Bergen. Boixaderas, M. R. 2012. Plant dispersal in a changing climate: a seed-rain study along climate gradients in southern Norway. Master thesis, University of Bergen, Bergen. Brooker, R. W. 2006. Plant-plant interactions and environmental change. New Phytologist 171:271-284. Brooker, R. W. and T. V. Callaghan. 1998. The balance between positive and negative plant interactions and its relationship to environmental gradients: a model. Oikos 81:196207. Callaway, R. M., R. W. Brooker, P. Choler, Z. Kikvidze, C. J. Lortie, R. Michalet, L. Paolini, F. I. Pugnaire, B. Newingham, E. T. Aschehoug, C. Armas, D. Kikodze, and B. J. Cook. 2002. Positive interactions among alpine plants increase with stress. Nature 417:844-848. Chapin III, F. S., B. H. Walker, R. J. Hobbs, D. U. Hooper, J. H. Lawton, O. E. Sala, and D. Tilman. 1997. Biotic control over the functioning of ecosystems. Science 277:500504. Choler, P., R. Michalet, and R. M. Callaway. 2001. Facilitation and competition on gradients in alpine plant communities. Ecology 82:3295-3308. 22

Cleland, E. E., I. Chuine, A. Menzel, H. A. Mooney, and M. D. Schwartz. 2007. Shifting plant phenology in response to global change. Trends in Ecology and Evolution 22:357-365. Cornelissen, J. H. C., T. V. Callaghan, J. M. Alatalo, A. Michelsen, E. Graglia, A. E. Hartley, D. S. Hik, P. Horswill, M. C. Press, C. H. Robinson, G. H. R. Henry, G. R. Shaver, G. K. Phoenix, D. G. Jones, S. Jonasson, F. S. Chapin III, U. Molau, C. Neill, J. A. Lee, J. M. Melillo, B. Sveinbjörnsson, and R. Aerts. 2001. Global change and arctic ecosystems: is lichen decline a function of increases in vascular plant biomass? Journal of Ecology 89:984-994. De Frenne, P. 2014. Innovative empirical approaches for inferring climate-warming impacts on plants in remote areas. New Phytologist. DOI: 10.1111/nph.12992. Dunne, J. A., J. Harte, and K. J. Taylor. 2003. Supalpine meadow flowering phenology responses to climate change: integrating experimental and gradient methods. Ecological Monographs 73:69-86. Dunne, J. A., S. R. Saleska, M. L. Fischer, and J. Harte. 2004. Integrating experimental and gradient methods in ecological climate change research. Ecology 85:904-916. Dunnett, N. P. and J. P. Grime. 1999. Competition as an amplifier of short-term vegetation responses to climate: an experimental test. Functional Ecology 13:388-395. Elmendorf, S. C., G. H. R. Henry, R. D. Hollister, R. G. Björk, A. D. Bjorkman, T. V. Callaghan, L. S. Collier, E. J. Cooper, J. H. C. Cornelissen, T. A. Day, A. M. Fosaa, W. A. Gould, J. Grétarsdóttir, J. Harte, L. Hermanutz, D. S. Hik, A. Hofgaard, F. Jarrad, I. S. Jónsdóttir, F. Keuper, K. Klanderud, J. A. Klein, S. Koh, G. Kudo, S. I. Lang, V. Loewen, J. L. May, J. Mercado, A. Michelsen, U. Molau, I. H. Myers-Smith, S. F. Oberbauer, S. Pieper, E. Post, C. Rixen, C. H. Robinson, N. M. Schmidt, G. R. Shaver, A. Stenström, A. Tolvanen, Ø. Totland, T. Troxler, C.-H. Wahren, P. J. Webber, J. M. Welker, and P. A. Wookey. 2012a. Global assessment of experimental climate warming on tundra vegetation: heterogeneity over space and time. Ecology Letters 15:164-175. Elmendorf, S. C., G. H. R. Henry, R. D. Hollister, R. G. Björk, N. Boulanger-Lapointe, E. J. Cooper, J. H. C. Cornelissen, T. A. Day, E. Dorrepaal, T. G. Elumeeva, M. Gill, W. A. Gould, J. Harte, D. S. Hik, A. Hofgaard, D. R. Johnson, J. F. Johnstone, I. S. Jónsdóttir, J. C. Jorgenson, K. Klanderud, J. A. Klein, S. Koh, G. Kudo, M. Lara, E. Lévesque, B. Magnússon, J. L. May, J. A. Mercado-Díaz, A. Michelsen, U. Molau, I. H. Myers-Smith, S. F. Oberbauer, V. G. Onipchenko, C. Rixen, N. M. Schmidt, G. R. Shaver, M. J. Spasojevic, Þ. E. Þórhallsdóttir, A. Tolvanen, T. Troxler, C. E. Tweedie, 23

S. Villareal, C.-H. Wahren, X. Walker, P. J. Webber, J. M. Welker, and S. Wipf. 2012b. Plot-scale evidence of tundra vegetation change and links to recent summer warming. Nature Climate Change 2:453-457. Fariñas, S. A. 2011. How do changing climate variables impact alpine plant communities? Linking gradients of temperature, precipitation, and available soil nitrogen to plant growth and chemistry. Master thesis, University of Michigan, Ann Arbor. Gellesch, E., R. Hein, A. Jaeschke, C. Beierkuhnlein, and A. Jentsch. 2013. Biotic interactions in the face of climate change. Progress in Botany 74:321-349. Gillespie, M. A. K., I. S. Jónsdóttir, I. D. Hodkinson, and E. J. Cooper. 2013. Aphid-willow interactions in a high Arctic ecosystem: responses to raised temperature and goose disturbance. Global Change Biology 19:3698-3708. Gilman, S. E., M. C. Urban, J. Tewksbury, G. W. Gilchrist, and D. R. Holt. 2010. A framework for community interactions under climate change. Trends in Ecology and Evolution 25:325-331. Grabherr, G., M. Gottfried, and H. Pauli. 1994. Climate effects on mountain plants. Nature 369:448. Grime, J. P. 2001. Plant strategies, vegetation processes, and ecosystem properties. Second edition. John Wiley & Sons, Ltd., Chichester. Grytnes, J.-A., J. Kapfer, G. Jurasinski, H. H. Birks, H. Henriksen, K. Klanderud, A. Odland, M. Ohlson, S. Wipf, and H. J. B. Birks. 2014. Identifying the driving factors behind observed elevational range shifts on European mountains. Global Ecology and Biogeography 23:876-884. Harsch, M. A. and J. HilleRisLambers. 2014. Species distributions shift downwards across western North America. Global Change Biology. DOI: 10.1111/gcb.12697. Harsch, M. A., P. E. Hulme, M. S. McGlone, and R. P. Duncan. 2009. Are treelines advancing? A global meta-analysis of treeline response to climate warming. Ecology Letters 12:1040-1049. He, Q., M. D. Bertness, and A. H. Altieri. 2013. Global shifts towards positive species interactions with increasing environmental stress. Ecology Letters 16:695-706. Hegland, S. J., A. Nielsen, A. Lázaro, A.-L. Bjerknes, and Ø. Totland. 2009. How does climate warming affect plant-pollinator interactions? Ecology Letters 12:184-195. Holzapfel, C., K. Tielbörger, H. A. Parag, J. Kigel, and M. Sternberg. 2006. Annual plantshrub interactions along an aridity gradient. Basic and Applied Ecology 7:268-279.

24

Hågvar, S. and K. Klanderud. 2009. Effects of simulated environmental change on alpine soil arthropods. Global Change Biology 15:2972-2980. IPCC. 2013. Climate change 2013: the physical science basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge. Kareiva, P. M., J. G. Kingsolver, and R. B. Huey. 1993. Introduction. In P. M. Kareiva, J. G. Kingsolver, and R. B. Huey (eds) Biotic interactions and global change. Sinauer Associates Inc. Publishers, Sunderland. Kausrud, K. L., A. Mysterud, H. Steen, J. O. Vik, E. Østbye, B. Cazelles, E. Framstad, A. M. Eikeset, I. Mysterud, T. Solhøy, and N. C. Stenseth. 2008. Linking climate change to lemming cycles. Nature 456:93-97. Kikvidze, Z., F. I. Pugnaire, R. W. Brooker, P. Choler, C. J. Lortie, R. Michalet, and R. M. Callaway. 2005. Linking patterns and processes in alpine plant communities: a global study. Ecology 86:1395-1400. Klanderud, K. 2005. Climate change effects on species interactions in an alpine plant community. Journal of Ecology 93:127-137. Klanderud, K. 2008. Species-specific responses of an alpine plant community under simulated environmental change. Journal of Vegetation Science 19:363-372. Klanderud, K. 2010. Species recruitment in alpine plant communities: the role of species interactions and productivity. Journal of Ecology 98:1128-1133. Klanderud, K. and Ø. Totland. 2005a. The relative importance of neighbours and abiotic environmental conditions for population dynamic parameters of two alpine plant species. Journal of Ecology 93:493-501. Klanderud, K. and Ø. Totland. 2005b. Simulated climate change altered dominance hierarchies and diversity of an alpine biodiversity hotspot. Ecology 86:2047-2054. Klanderud, K. and Ø. Totland. 2007. The relative role of dispersal and local interactions for alpine plant community diversity under simulated climate warming. Oikos 116:12791288. Klanderud, K. and Ø. Totland. 2008. Diversity-stability relationships of an alpine plant community under simulated environmental change. Arctic, Antarctic, and Alpine Research 40:679-684. Lang, S. I., J. H. C. Cornelissen, G. R. Shaver, M. Ahrens, T. V. Callaghan, U. Molau, C. J. F. ter Braak, A. Hölzer, and R. Aerts. 2012. Arctic warming on two continents has

25

consistent negative effects on lichen diversity and mixed effects on bryophyte diversity. Global Change Biology 18:1096-1107. Lavergne, S., N. Mouquet, W. Thuiller, and O. Ronce. 2010. Biodiversity and climate change: integrating evolutionary and ecological responses of species and communities. Annual Review of Ecology, Evolution and Systematics 41:321-350. Lenoir, J., J. C. Gégout, P. A. Marquet, P. de Ruffray, and H. Brisse. 2008. A significant upward shift in plant species optimum elevation during the 20th century. Science 320:1768-1771. Luo, Y., D. Gerten, G. Le Maire, W. J. Parton, E. Weng, X. Zhou, C. Keough, C. Beier, P. Ciais, W. Cramer, J. S. Dukes, B. Emmett, P. J. Hanson, A. Knapp, S. Linder, D. Nepstad, and L. Rustad. 2008. Modeled interactive effects of precipitation, temperature, and [CO2] on ecosystem carbon and water dynamics in different climatic zones. Global Change Biology 14:1986-1999. Maestre, F. T., F. Valladares, and J. F. Reynolds. 2005. Is the change of plant-plant interactions with abiotic stress predictable? A meta-analysis of field results in arid environments. Journal of Ecology 93:748-757. Marion, G. M., G. H. R. Henry, D. W. Freckman, J. Johnstone, G. Jones, M. H. Jones, E. Lévesque, U. Molau, P. Mølgaard, A. N. Parsons, J. Svoboda, and R. A. Virginia. 1997. Open-top designs for manipulating field temperature in high-latitude ecosystems. Global Change Biology 3:20-32. Meineri, E., O. Skarpaas, J. Spindelböck, T. Bargmann, and V. Vandvik. 2014. Direct and size-dependent effects of climate on flowering performance in alpine and lowland herbaceous species. Journal of Vegetation Science 25:275-286. Meineri, E., J. Spindelböck, and V. Vandvik. 2013. Seedling emergence responds to both seed source and recruitment site climates: a climate change experiment combining transplant and gradient approaches. Plant Ecology 214:607-619. Memmott, J., P. G. Craze, N. M. Waser, and M. V. Price. 2007. Global warming and the disruption of plant-pollinator interactions. Ecology Letters 10:710-717. Michalet, R., C. Schöb, C. J. Lortie, R. W. Brooker, and R. M. Callaway. 2014. Partitioning net interactions among plants along altitudinal gradients to study community responses to climate change. Functional Ecology 28:75-86. Moen, A. 1998. Nasjonalatlas for Norge: vegetasjon. Statens Kartverk, Hønefoss. Montoya, J. M. and D. Raffaelli. 2010. Climate change, biotic interactions and ecosystem services. Philosophical Transactions of the Royal Society B 365:2013-2018. 26

Mooney, H., A. Larigauderie, M. Cesario, T. Elmquist, O. Hoegh-Guldberg, S. Lavorel, G. M. Mace, M. Palmer, R. Scholes, and T. Yahara. 2009. Biodiversity, climate change, and ecosystem services. Environmental Sustainability 1:46-54. Norwegian Meteorological Institute. 2010. eKlima. Data from weather station 25840 at Finse, normal period 1961-1990. To be found on the web page eklima.met.no. Olofsson, J., L. Oksanen, T. Callaghan, P. E. Hulme, T. Oksanen, and O. Suominen. 2009. Herbivores inhibit climate-driven shrub expansion on the tundra. Global Change Biology 15:2681-2693. Pace, M. L. 1993. Forecasting ecological responses to global change: the need for large-scale comparative studies. In P. M. Kareiva, J. G. Kingsolver, and R. B. Huey (eds) Biotic interactions and global change. Sinauer Associates Inc. Publishers, Sunderland. Parmesan, C. 2006. Ecological and evolutionary responses to recent climate change. Annual Review of Ecology, Evolution and Systematics 37:637-669. Parmesan, C. and G. Yohe. 2003. A globally coherent fingerprint of climate change impacts across natural systems. Nature 421:37-42. Pauli, H., M. Gottfried, S. Dullinger, O. Abdaladze, M. Akhalkatsi, J. L. B. Alonso, G. Coldea, J. Dick, B. Erschbamer, R. F. Calzado, D. Ghosn, J. I. Holten, R. Kanka, G. Kazakis, J. Kollár, P. Larsson, P. Moiseev, D. Moiseev, U. Molau, J. M. Mesa, L. Nagy, G. Pelino, M. Puşcaş, G. Rossi, A. Stanisci, A. O. Syverhuset, J.-P. Theurillat, M. Tomaselli, P. Unterluggauer, L. Villar, P. Vittoz, and G. Grabherr. 2012. Recent plant diversity changes on Europe's mountain summits. Science 336:353-355. Post, E. 2013. Ecology of climate change: the importance of biotic interactions. Princeton University Press, Princeton. Post, E. and C. Pedersen. 2008. Opposing plant community responses to warming with and without herbivores. Proceedings of the National Academy of Sciences 105:1235312358. Pugnaire, F. I. and M. T. Luque. 2001. Changes in plant interactions along a gradient of environmental stress. Oikos 93:42-49. Pötsch, C. 2010. Resource allocation in two selected Carex species along climate gradients. Master thesis, University of Bergen, Bergen. Ravolainen, V. T., K. A. Bråthen, N. G. Yoccoz, J. K. Nguyen, and R. A. Ims. 2014. Complementary impacts of small rodents and semi-domesticated ungulates limit tall shrub expansion in the tundra. Journal of Applied Ecology 51:234-241.

27

Root, T. L., J. T. Price, K. R. Hall, S. H. Schneider, C. Rosenzweig, and J. A. Pounds. 2003. Fingerprints of global warming on wild animals and plants. Nature 421:57-60. Rustad, L. E., J. L. Campbell, G. M. Marion, R. J. Norby, M. J. Mitchell, A. E. Hartley, J. H. C. Cornelissen, J. Gurevitch, and GCTE-NEWS. 2001. A meta-analyses of the response of soil respiration, net nitrogen mineralization, and aboveground plant growth to experimental ecosystem warming. Oecologia 126:543-562. Sala, O. E., F. S. Chapin III, J. J. Armesto, E. Berlow, J. Bloomfield, R. Dirzo, E. HuberSanwald, L. F. Huenneke, R. B. Jackson, A. Kinzig, R. Leemans, D. M. Lodge, H. A. Mooney, M. Oesterheld, N. L. Poff, M. T. Sykes, B. H. Walker, M. Walker, and D. H. Wall. 2000. Global biodiversity scenarios for the year 2100. Science 287:1770-1774. Soliveres, S. and F. T. Maestre. 2014. Plant-plant interactions, environmental gradients and plant diversity: a global synthesis of community-level studies. Perspectives in Plant Ecology, Evolution and Systematics 16:154-163. Soliveres, S., C. Smit, and F. T. Maestre. 2014. Moving forward on facilitation research: response to changing environments and effects on the diversity, functioning and evolution of plant communities. Biological Reviews. DOI: 10.1111/brv.12110. Sturm, M., C. Racine, and K. Tape. 2001. Increasing shrub abundance in the Arctic. Nature 411:546-547. Suttle, K. B., M. A. Thomsen, and M. E. Power. 2007. Species interactions reverse grassland responses to changing climate. Science 315:640-642. Thomas, C. D., A. M. A. Franco, and J. K. Hill. 2006. Range retractions and extinctions in the face of climate warming. Trends in Ecology and Evolution 21:415-416. Thuiller, W. 2007. Climate change and the ecologist. Nature 448:550-552. Tingstad, L. 2012. Tree seedling establishment along climatic gradients from lowland to alpine areas in western Norway. Master thesis, Norwegian University of Life Sciences, Ås. Tylianakis, J. M., R. K. Didham, J. Bascompte, and D. A. Wardle. 2008. Global change and species interactions in terrestrial ecosystems. Ecology Letters 11:1351-1363. Urban, M. C., J. J. Tewksbury, and K. S. Sheldon. 2012. On a collision course: competition and dispersal differences create no-analogue communities and cause extinction during climate change. Proceedings of the Royal Society B 279:2072-2080. van der Putten, W., M. Macel, and M. E. Visser. 2010. Predicting species distribution and abundance responses to climate change: why it is essential to include biotic

28

interactions across trophic levels. Philosophical Transactions of the Royal Society B 365:2025-2034. Visser, M. E. and C. Both. 2005. Shifts in phenology due to global climate change: the need for a yardstick. Proceedings of the Royal Society B 272:2561-2569. Walker, M. D., C. H. Wahren, R. D. Hollister, G. H. R. Henry, L. E. Ahlquist, J. M. Alatalo, M. S. Bret-Harte, M. P. Calef, T. V. Callaghan, A. B. Carroll, H. E. Epstein, I. S. Jónsdottir, J. A. Klein, B. Magnússon, U. Molau, S. F. Oberbauer, S. P. Rewa, C. H. Robinson, G. R. Shaver, K. N. Suding, C. C. Thompson, A. Tolvanen, Ø. Totland, P. L. Turner, C. E. Tweedie, P. J. Webber, and P. A. Wookey. 2006. Plant community responses to experimental warming across the tundra biome. Proceedings of the National Academy of Sciences 103:1342-1346. Walther, G.-R., E. Post, P. Convey, A. Menzel, C. Parmesan, T. J. C. Beebee, J.-M. Fromentin, O. Hoegh-Guldberg, and F. Bairlein. 2002. Ecological responses to recent climate change. Nature 416:389-395. Watkinson, A. R. and J. C. Powell. 1993. Seedling recruitment and the maintenance of clonal diversity in plant populations - a computer simulation of Ranunculus repens. Journal of Ecology 81:707-717.

29

Paper I

The role of temperature, precipitation and competition as determinants of tree seedling recruitment across the tree line ecotone

Lise Tingstad1,2*, Siri L. Olsen3, Kari Klanderud3, Vigdis Vandvik1 and Mikael Ohlson3

1

Department of Biology, University of Bergen, 5020 Bergen, Norway

2

Norwegian Forest and Landscape Institute, 5244 Fana, Norway

3

Department of Ecology and Natural Resource Management, Norwegian University of Life

Sciences, 1432 Ås, Norway

* Corresponding author: Lise Tingstad [email protected]

Author contributions: MO conceived and planned the experiment, VV and KK designed the climate grid, set up the field localities and designed the field experiment, LT performed the field work, SLO, LT, KK and MO analyzed the data, SLO and LT wrote the manuscript. All authors provided extended editorial advice. 1

Abstract Seedling recruitment is a critical life history stage in conifer trees. Successful emergence and establishment are both tightly linked to climate and biotic interactions. However, these interactions are complex, and more knowledge is needed on the combined effects of climate and biotic interactions on tree seedling recruitment. We conducted a seed sowing experiment to investigate how temperature, precipitation and biotic interactions impact emergence and establishment of Scots pine (Pinus sylvestris) and Norway spruce (Picea abies) seedlings above and below the tree line along natural climatic gradients in southern Norway. Seeds were sown in intact vegetation and in experimentally created gaps where the surrounding vegetation had been removed. Interestingly, no temperature limitation was detected on seedling emergence and early establishment in this study. Contrary to our expectations, significantly higher numbers of seedlings established at low temperatures, and very few seedlings were detected at the warmer lowland localities. Further we found high emergence and establishment in gap plots, even at lower temperature sites. Our results suggest that biotic interactions may override temperature as the main limiting factor for tree seedling recruitment in the sub- and low-alpine zones of southern Norway. We show that successful emergence and establishment of trees depend on a complex interplay between climate, including both temperature and precipitation, and biotic interactions.

Key words: tree seedling recruitment - biotic interactions - natural climate gradients

2

Introduction All populations are dependent on successful recruitment for their long-term persistence. Recruitment is consequently a key life history event, and for trees it involves several phases and processes, e.g. flowering, pollination, seed maturation, seed dispersal, seed germination, early seedling establishment and survival of seedlings, all of which are closely linked to climatic conditions and biotic interactions (Bansal and Germino 2010; Grubb 1977; Kitajiama and Fenner 2000; Turnbull et al. 2000). Under harsh environmental conditions, as in the alpine tree-line ecotone, it can take a long time for a tree seed to establish and develop into a young tree. The specific biotic and abiotic requirements of the seedlings must thus be fulfilled and prevail for several seasons to secure successful tree recruitment (Juntunen and Neuvonen 2006). Failure to meet one or more of the requirements throughout the seedling stage is the main reason why this life-stage typically has the highest mortality rate (Germino et al. 2002). Knowledge of the relative importance of biotic and abiotic factors affecting early life-stages is thus paramount to understanding the recruitment dynamics of tree species in space and time.

Abiotic factors are often considered the most important determinants of plant recruitment in climatically harsh and cold environments. Processes prior to seedling establishment, as well as seedling establishment itself, are heavily impacted by temperature and precipitation (McCarty 2001; Smith 1994). In boreal conifer trees (e.g. Pinus and Picea species), germination typically peaks at temperatures slightly above 20 °C, while periods with temperatures below 15 °C strongly limit seed germination (Black and Bliss 1980; Mork 1938). Increasing temperature is in general assumed to favour plant recruitment in lowtemperature environments like the alpine (Fenner and Thompson 2005). Periods of drought also limit seed germination severely, and precipitation is known to have a direct positive effect on seedling establishment in dry environments (McCarty 2001; Walther et al. 2002). However, the effects of precipitation on recruitment are hard to predict as they depend on the amount, the timing and the reliability of the rainfall (Fay and Schultz 2009). While the seedlings’ abiotic requirements must be fulfilled to secure successful recruitment, biotic interactions also play a key role in tree seedling emergence, establishment and survival (Grubb 1977; Hörnberg et al. 1997; Ohlson and Zackrisson 1992). A recent study on alpine plants indicates that once a plant seedling has germinated in a suitable microhabitat, biotic interactions become one of the major drivers of the long term establishment success (Olsen 3

and Klanderud 2014). However, effects of biotic interactions are known to vary with abiotic conditions. According to the stress-gradient hypothesis (Brooker and Callaghan 1998), the net outcome of biotic interactions differ along gradients of abiotic environmental stress, where facilitation and competition dominate at high and low stress levels, respectively (Callaway et al. 2002). Hence, competition should be more important as a determining recruitment factor in habitats characterized by relatively warm and wet conditions as compared to the situation in habitats that are relatively cold and dry, where facilitative interactions likely prevail.

In this study we investigate the interactive effects of climate and biotic interactions for the recruitment of two common boreal tree species, Scots pine (Pinus sylvestris L.) and Norway spruce (Picea abies (L.) Karst.), by sowing seeds in intact vegetation and in bare-ground gaps across a unique ‘climate grid’ in which a natural temperature gradient (alpine-lowland) is repeated over four levels of precipitation (see Meineri et al. 2013, 2014). Scots pine and Norway spruce are dominant forest tree species within the boreal zone of Eurasia. Both species are native to Norway and known for their wide ecological range (Ohlson and Zackrisson 1992; Seppä et al. 2009). The climate grid encompasses the dynamic tree-line ecotone, in which upward migration of trees has been studied extensively in a climate warming perspective (Holtmeier and Broll 2005; Kozlowski et al. 1991; Kullman 2002; Körner and Paulsen 2004). Despite the comprehensive studies of tree lines, we still lack knowledge about the relative importance of biotic and climatic factors as determinants of tree recruitment in alpine and subalpine environments. Understanding the factors impacting successful recruitment is a fundamental step towards understanding tree-line dynamics, especially in a changing climate.

We have addressed the following questions: 1) how do emergence, establishment and biomass allocation in pine and spruce seedlings vary with temperature and precipitation? 2) how do biotic interactions impact seedling recruitment and biomass allocation?, and 3) how do the effect of biotic interactions vary with climate? We hypothesized that the effect of gap disturbance (i.e. release from biotic interactions) on seedling emergence and establishment would become increasingly positive with increasing temperature, in accordance with the stress-gradient hypothesis. For cold sites we expected to find negative effects of gaps (i.e. facilitation), as neighbouring plants may provide shelter from harsh abiotic conditions such as wind, frost and drought. In contrast, we expected positive effects of gaps (i.e. release from competition) in warm sites with relatively favourable abiotic conditions. As climatic 4

conditions are likely to have significant impact on plant allocation (Skarpaas et al. submitted), we also examined the growth and allocation patterns of the tree seedlings across the same climate grid. Our expectations are in line with the optimal partitioning theory, stating that plants should allocate biomass to organs that acquire the most limiting resource (McCarthy and Enquist 2007).

Methods Study area, study species and experimental design The seed sowing experiment was conducted within a unique climate grid consisting of a natural temperature gradient replicated four times along a precipitation gradient stretching from the oceanic coast toward the more continental inland in southern Norway (Fig. 1) (see also Meineri et al. 2013, 2014). The design of the grid allows for the independent combination of three levels of summer temperature [means of the four warmest months: 6.5°C, (alpine), 8.5°C (intermediate) and 10.5°C (lowland) ] with four levels of mean annual precipitation [600, 1200, 2000 and 2700 mm] across twelve sites (Fig. 1 and Table 1).

The sites were all established in semi-natural grasslands on calcareous bedrock (Berge 2010; Meineri et al. 2013, 2014), supporting high fine-scale plant diversity. The four lowland sites were situated in or near forested areas; the four intermediate sites were situated closer to the tree line, and the four alpine sites were situated above the tree line in the low-alpine zone. Besides their climatic attributes, sites were selected specifically to keep vegetation type, geology and land-use as constant as possible to facilitate comparison among sites. Further details on vegetation characteristics and site selection criteria are described in Meineri et al. (2013, 2014).

Each of the twelve sites was split into 5 blocks, with four study plots of 25 × 25 cm randomly positioned within each block. There were one control plot (plot with intact vegetation) and one gap plot (plots where the upper soil layer had been removed experimentally to simulate small-scale disturbance) per species in each block. In total there were 240 study plots; 120 for each species.

5

All seeds used in the experiment were bought from “Skogfrøverket” in Lillehammer, Norway, and were stored at -20 °C prior to sowing. In spring 2010, 50 seeds of each species were sown in each plot. The seeds were scattered evenly on the bare soil of the gap plots and in the intact vegetation. To prevent seeds from being blown or washed away from the gap plots, seeds were carefully pressed down a few millimetres into soil surface, but not covered. The sites were fenced to prevent animal trampling and disturbance, and the grassland sward was cut approximately 5 cm above ground level in late August to mimic the biomass removal by the abundant free-roaming domestic or wild ungulates in the area. Otherwise seedlings were left undisturbed.

Seedling recruitment was assessed by recording emergence, survival and growth of all seedlings twice during the early establishment phase. All seedlings were counted 5, 12 and 16 months after sowing. After 16 months (end of second growing season) all seedlings were harvested. In gap plots with three or more seedlings, three seedlings were harvested with roots carefully excavated. The remaining seedlings were harvested above ground. In gap plots with less than three seedlings and in intact vegetation plots, all seedlings were harvested above ground. Post-harvest, seedlings were stored in paper bags at 4°C and no light until further processing could take place. The harvested seedlings were dried at 80°C until constant weight. The height (total length from the original emerging point to the apical meristem), aboveground biomass, root length and below-ground biomass of seedlings were measured individually.

Statistical analyses The effects of climatic variables and disturbance on tree seedling emergence, survival and growth were examined using linear mixed-effects models. The lowland sites were omitted from these analyses as almost no seedlings emerged there (Fig. 2). Temperature, precipitation and treatment (gap versus intact vegetation) were used as fixed factors, and precipitation was given as a categorical value (1-4). To account for the nested design, we estimated random intercepts for blocks nested in sites. We assumed Poisson distributions for seedling emergence and establishment since these are count data and do not fit a normal distribution. Normal distributions were assumed for seedling height and weight, and here likelihood ratio (LR) tests were used to select the final models, with maximum likelihood for the LR-tests and 6

restricted maximum likelihood for estimating model coefficients. For the latter, Markov Chain Monte Carlo (MCMC) estimation with 10 000 iterations was used to assess variable significances. Separate mixed-effects models were run for each tree species. All statistical analyses were performed in R version 2.15.2, (R Development Core Team, 2012) using R Studio Version 0.96.331 (RStudio, Inc.). We used the package lme4 (Bates et al. 2012) for the mixed effects models and LanguageR (Baayen 2011) for the Markov Chain Monte Carlo estimations.

Results In total, 1226 pine seedlings and 1138 spruce seedlings were recorded in autumn 2010 and recounted spring 2011. Due to very low over-winter mortality (< 5 %), we used the recordings from spring 2011 as counts of emergence. In autumn 2011, after the second growing season, 900 pine seedlings and 922 spruce seedlings were recorded (the establishment count) and harvested. The numbers of emerged seedlings per plot ranged from 0 to 41 for pine, and from 0 to 44 for spruce, with a rounded median of 9 and 10, respectively. Overall, patterns in seedling emergence, establishment and growth followed similar trends for both pine and spruce.

Emergence rates within the closed vegetation in alpine sites were higher than in intermediate sites, while there were almost no emerged seedlings in lowland sites (Fig. 2, Table 2), where only a few spruce seedlings and no pine seedlings were recorded. The effect of precipitation on seedling emergence in intact vegetation was stronger in intermediate than in alpine sites and was especially pronounced for spruce (Table 2). In general, more seedlings were found in the wet western part of the grid, with a pronounced peak at precipitation level 3 for spruce and precipitation level 4 for pine. This pattern was found for both intermediate and alpine sites. Vegetation removal favoured conifer seedling emergence, as significantly more seedlings of both species were found in gap plots compared to intact vegetation in both intermediate and alpine sites. The gap effect was strongest in intermediate sites and varied significantly with precipitation, with the strongest gap effect at intermediate to high precipitation levels (Table 2). Gap plots in sites with intermediate temperature and medium high precipitation (level 3) supported the highest number of seedlings per plot for both species.

7

Establishment rates of both study species were higher in alpine sites than in intermediate sites, although significant for pine seedlings only (Table 2). In lowland sites the numbers were low, as expected from the low emergence rate (Fig. 2). However, a few “late emergents” in lowland sites were detected in the third count, and were included in the establishment counts. The establishment rate was higher at intermediate precipitation levels at both intermediate and alpine sites, and the relationship between establishment and precipitation followed the same trends as those for emergence rate and precipitation (Fig. 2), but with fewer significant interactions. The gap treatment had a positive effect on establishment, with more seedlings establishing in gap plots compared to intact vegetation in both intermediate and alpine sites (Fig. 2 and Table 2).

Both pine and spruce seedling height varied along the temperature and precipitation gradients, with a tendency towards taller seedlings in sites with intermediate temperature and precipitation (Fig. 3, Table 2). The few pine seedlings (the late emergents) found in the lowland, were among the tallest seedlings registered (Fig. 2). Seedlings were generally taller in intact vegetation than in gaps (Fig. 3, Table 2). At intermediate temperatures, the height difference of seedlings in gaps versus intact plots decreased with increasing precipitation levels (Fig. 3). Aboveground biomass was higher for seedlings from intermediate sites than from alpine sites. For pine seedlings we found a trend of decreasing biomass with increasing precipitation in the alpine. Seedlings of both species had a lower biomass in intact vegetation than in gaps (Fig. 3). The gap effect on pine seedling biomass was significantly stronger in intermediate sites than in alpine sites (Table 2).

In general, roots of both species were found to be relatively longer in alpine sites, and decreased in length and biomass towards warmer climates in intermediate sites (not shown).

Discussion Contrary to our expectations, low temperatures did not restrict emergence or establishment of pine and spruce seedlings under the current temperature regime. We found that recruitment of tree seedlings was significantly higher in alpine sites compared to the warmer intermediate sites. This was unexpected, as the alpine sites are located well above the tree line. Further, we found extremely few seedlings in lowland sites, which was also an unexpected result as all 8

lowland sites are located in a vegetation zone characterized by well-developed forests and should in principle provide suitable microhabitats for tree seedlings. Our findings contradict previous studies showing a dominant role of temperature for successful establishment of trees (Grace et al. 2002; Juntunen and Neuvonen 2006; Kullman and Engelmark 1990; Payette 1985; Tranquillini 1979) and suggest that other factors may override low temperature as the main determinant of successful tree seedling emergence and establishment.

Within our temperature range, precipitation was an important factor affecting tree seedling emergence. Thus, our findings are in line with numerous studies relating recruitment to soil moisture and showing its importance for initiating germination mechanisms (e.g. Ibanez et al. 2007). The tendency for a unimodal relationship with a decline in establishment rates in sites with the lowest and highest amount of precipitation may be caused partially by drought in the dry end of the precipitation gradient, and by hypoxic soil conditions due to water-saturation at higher precipitation levels. Seedling survival at medium-high precipitation levels may also be enhanced by high winter precipitation. Frost heaving has been reported to be among the most common causes of tree seedling damage and mortality during the first winter (Erefur et al. 2008), stressing the importance of an insulating snow cover. Previous studies found little evidence that winter conditions affect tree seedling survival patterns (Hadley and Smith 1986), and high over-winter survival of seedlings found in this study may relate to the seedlings’ low height and snow coverage during winter.

While both precipitation and temperature were found to have an effect on tree seedling recruitment success, biotic interactions also affected recruitment across the ecotone. Gaps generally promoted seedling emergence and establishment in both intermediate and alpine sites, as more seedlings were found in gaps compared to intact vegetation. Our findings are in line with previous studies showing higher emergence and establishment of tree seedlings in gaps than intact vegetation (Berkowitz et al. 1995; Gray and Spies 1996; Munier et al. 2010). Thus, it seems that decreased competition, with a corresponding increase in light availability due to removal of neighbours, overrides the potential negative effects of gaps such as increased vulnerability to injuries and damage (Munier et al. 2010) and drought- and temperature-related stress (Smith et al. 2003). According to the stress-gradient hypothesis the role of competition should decrease relative to facilitation with increasing abiotic stress, and facilitation should therefore be more important in alpine than in lowland habitats (Callaway et al. 2002). We found that the positive effect of lower competition in gap plots was relatively 9

more important than the facilitative effect of surrounding vegetation at all temperatures, including in the alpine. Nonetheless, the intensity of competition from the surrounding vegetation seemed to increase with increasing temperature. Our findings suggest that competition is an important determinant of successful tree seedling emergence and establishment of pine and spruce seedlings in the tree-line ecotone, although the strength of the interactions varies with temperature and precipitation.

A clear indication of competition was that seedlings in intact vegetation were taller than seedlings in gap plots, but had a comparably lower aboveground biomass. In intact vegetation with intense competition for light seedlings are forced to invest relatively more in height growth, potentially at the expense of stability tissue like thicker stem and roots (Norgren 1996). The difference in biomass between pine seedlings in gaps and in intact vegetation was larger in intermediate compared to alpine sites, again indicating that competition intensity increased with increasing temperature. The plots with intact vegetation had quite dense stands of graminoids and forbs, especially in lowland sites where the vegetation height can reach up to 21 cm (unpublished data), making it hard for seedlings to compete for space, light and nutrients. Even gaps were overgrown in some sites (L. Tingstad pers. obs.) which might provide some explanation for the almost complete lack of tree seedling recruitment in the lowlands.

Although low temperature was not a limiting factor for the emergence and establishment of conifer seedlings in this study, it may limit growth and survival of trees at later life stages. Accordingly, previous studies have detected tree seedlings above the tree line that rarely become mature, upright trees (Körner 2012; Körner and Paulsen 2004). Seedlings in our study had a mean height of 2.45 cm and may have experienced a thermal advantage of being short and sheltered by surrounding vegetation. The apical meristem of a taller tree experiences a temperature different from the microclimate near ground, and taller plants are more prone to abiotic factors like wind and ice pinning (Smith et al. 2003). If low temperature currently limits survival and growth of trees beyond the seedling stage, climate warming may reduce this effect, which may in turn lead to altitudinal advancement of trees. However, climate warming is also expected to increase the role of competition in plant communities (Klanderud 2005; Olsen and Klanderud 2014), thereby further limiting tree seedling emergence and establishment. Disturbance, i.e. the formation of gaps, may thus become increasingly important for recruitment of trees, as shown by the positive gap effect found in this study. In 10

conclusion, future tree-line dynamics will be difficult to predict, as recruitment of new trees depends on a complex interplay between climate, including both temperature and precipitation, and biotic interactions.

Acknowledgements This project was funded by the Norwegian Research Council (NORKLIMA grant 184912/S30). We thank the land-owners for access to the field sites and the whole SeedClimteam for good collaboration and team-work in the field. We also thank anonymous reviewers for helpful comments on the manuscript.

11

References Baayen RH (2011) Data sets and functions with "Analyzing linguistic data: a practical introduction to statistics". Version 1.4.1 Bansal S, Germino MJ (2010) Variation in ecophysiological properties among conifers at an ecotonal boundary: comparison of establishing seedlings and established adults at timberline. Journal of Vegetation Science 21:133-142. doi: 10.1111/j.1654-1103.2009.01127.x Bates DM, Maechler M, Bolker B (2012) lme4: linear and mixed-effect models using S4 classes. R package verson 0.999999-0 Berge A (2010) Seedbank, seedrain and seedling recruitment along climate gradients in Southern Norway. Master thesis. Department of Biology, University of Bergen, Bergen Berkowitz AR, Canham CD, Kelly YVR (1995) Competition vs. facilitation of tree seedling growth and survival in early successional communities. Ecology 76:1156-1168 Black RA, Bliss LC (1980) Reproductive ecology of Picea mariana (Mill.) BSP., at tree line near Inuvik, Northwest Territories, Canada. Ecological Monographs 50:331-354. doi: 10.2307/2937255 Brooker RW, Callaghan TV (1998) The balance between positive and negative plant interactions and its relationship to environmental gradients: a model. Oikos:196-207 Callaway RM, Brooker RW, Choler P, Z. K, Lortie CJ, Michalet R, Paolini L, Pugnaire FI, Newingham B, Aschehoug ET, Armas C, Kikodze D, Cook BJ (2002) Positive interactions among alpine plants increase with stress. Nature 417: 844-848 Erefur C, Bergsten U, DeChantal M (2008) Establishment of direct seeded seedlings of Norway spruce and Scots pine: effects of stand and conditions, orientation and distance with respect to shelter tree, and fertilisation. Forest Ecology and Management 255:1186-1195 Fay PA, Schultz MJ (2009) Germination, survival and growth of grass and forb seedlings: effects of soil moisture and variability. Acta Oecologia 35:679-684 Fenner M, Thompson K (2005) The ecology of seeds. The Press Syndicate of the University of Cambridge, Cambridge Germino MJ, Smith WK, Resor AC (2002) Conifer seedling distribution and survival in an alpinetreeline ecotone. Plant Ecology 162:157-168 Grace J, Berninger F, Nagy L (2002) Impacts of climate change on the tree line. Annals of Botany 90:537-544 Gray N, Spies TA (1996) Gap size, within-gap position and canopy structure effects on conifer seedling establishment. Journal of Ecology 84:635-645. doi: 10.2307/2261327 Grubb PJ (1977) The maintenance of species-richness in plant communities: the importance of the regeneration niche Biol. Rev. Cambridge Philosophic. Soc. 52:107-145. doi: 10.1111/j.1469185X.1977.tb01347.x 12

Hadley JL, Smith WK (1986) Wind effects on needles of timberline conifers: seasonal influence on mortality. Ecology 67:12-19. doi: 10.2307/1938498 Holtmeier F-K, Broll G (2005) Sensitivity and response of northern hemisphere altitudinal and polar treelines to environmental change at landscape and local scales. Global Ecology and Biogeography 14:395-410. doi: 10.1111/j.1466-822X.2005.00168.x Hörnberg G, Ohlson M, Zackrisson O (1997) Influence of bryophytes and microrelief conditions on Picea abies seed regeneration patterns in boreal old-growth swamp forests. Canadian Journal of Forest Research 27:1015-1023. doi: 10.1139/x97-045 Ibanez I, Clark JS, LaDeau S, Lambers JHR (2007) Exploiting temporal variability to understand tree recruitment response to climate change. Ecological Monographs 77:163-177. doi: 10.1890/061097 Juntunen V, Neuvonen S (2006) Natural regeneration of Scots pine and Norway spruce close to the timberline in northern Finland. Silva Fennica 40:443 Kitajiama K, Fenner M (2000) The ecology of seedling generation In: Fenner M (ed), CAB International, Wallingford, pp 331-360 Klanderud K (2005) Climate change effects on species interactions in alpine plant communities. Journal of Ecology 93:127-137. doi: 10.1111/j.1365-2745.2004.00944.x Kozlowski TT, Kramer PJ, Pallardy SG (1991) The physiological ecology of woody plants. Academic Press, San Diego Kullman L (2002) Rapid recent range-margin rise of tree and shrub species in the Swedish Scandes. Journal of Ecology 90:68-77 Kullman L, Engelmark O (1990) A high Late Holocene tree-limit and the establisment of the spruce forest limit: a case study in Northern Sweden. Boreas 19:323-331 Körner C (2012) Tree lines will be understood once the functional difference between a tree and a shrub is. Ambio 41:197-206. doi: 10.1007/s13280-012-0313-2 Körner C, Paulsen J (2004) A world-wide study of high altitude treeline temperatures. Journal of Biogeography 31:713-732. doi: 10.1111/j.1365-2699.2003.01043.x McCarthy MC, Enquist BJ (2007) Consistency between an allometric approach and optimal partitioning theory in global patterns of plant biomass allocation. Functional Ecology 21:713720. doi: 10.1111/j.1365-2435.2007.01276.x McCarty JP (2001) Ecological consequences of recent climate change. Conservation Biology 15:320331 Meineri E, Spindelböck J, Vandvik V (2013) Seedling emergence responds to both seed source and recruitment site climates: a climate change experiment combining transplant and gradient approaches. Plant Ecology 214:607-619. doi: 10.1007/s11258-013-0193-y

13

Meineri E, Skarpaas O, Spindelböck J, Bargmann, T, Vandvik V (2014) Direct and size-dependent effect of climate on flowering performance in alpine and lowland herbaceous species. Journal of Vegetation Science 25:275-286 Mork E (1938) Gran og furu-frøets spiring ved forskjellige temperatur og fuktighet. Meddr. norske Skogfors.Ves. 6:225-249 Munier A, Hermanutz L, Jacobs J, Lewis K (2010) The interacting effects of temperature, ground disturbance and herbivory on seedling establishment: implications for the treeline advance with climate warming. Plant Ecology 210:19-30. doi: 10.1007/s11258-010-9724-y Norgren O (1996) Growth analysis of Scots pine and lodgepole pine seedlings. Forest Ecology and Management 86:15-26. doi: 10.1016/S0378-1127(96)03800-5 Ohlson M, Zackrisson O (1992) Tree establishment and microhabitat relationships in North Swedish peatlands. Canadian Journal of Forest Research 22:1869-1897. doi: 10.1139/x92-244 Olsen SL, Klanderud K (2014) Biotic interactions limit species richness in an alpine plant community,especially under experimentaly warming. Oikos 123:71-78. doi: 10.1111/j.16000706.2013.00336.x Payette SF, L. (1985) White spruce expansion at the tree line and recent climate change. Canadian Journal of Forest Research 15:241-251. doi: 10.1139/x85-042 R Development Core Team (2012) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna Seppä H, Alenius T, Bradshaw RHW, Gieschke T, Heikkilä M, Muukkonen P (2009) Invasion of Norway spruce (Picea abies) and the rise of the boreal ecosystem in Fennoscandia. Journal of Ecology 97:629-640. doi: 10.1111/j.1365-2745.2009.01505.x Smith RIL (1994) Vascular plants as bioindicators of regional warming in Antarctica. Oecologia 99:322-328. doi: 10.1007/BF00627745 Smith WK, Germino MJ, Hancock TE, Johnson DM (2003) Another perspective on altitudinal limits of alpine timberlines. Tree Physiology 23:1101-1112. doi: 10.1093/treephys/23.16.1101 Tranquillini W (1979) Physiological ecology of the alpine timberline: tree existence at high altitudes with special reference to the European Alps. Springer-Verlag, Berlin. Turnbull LA, Crawley MJ, Rees M (2000) Are plant populations seed-limited? A review of seed sowing experiments. Oikos 88:225-238. doi: 10.1034/j.1600-0706.2000.880201.x Tveito OE, Bjørdal I, Skjelvåg AO, Aune B (2005) A GIS-based agro-ecological decision system based on gridded climatology. Meteorological Applications 12:57-68. doi: doi:10.1017/S1350482705001490 Walther GR, Post E, Convey P, Menzel A, Parmesan C, Beebee TJC, Fromentin JM, Heogh-Guldberg IO, Bairlein F (2002) Ecological responses to climate change. Nature 416:389-395. doi: 10.1038/416389a

14

Table 1. The three elevation categories (lowland, intermediate and alpine), four precipitation categories (1-4) and climatic attributes of the twelve study sites. Tetra-term temperature (mean temperature of the four warmest months, °C) and precipitation (mean annual, mm) are provided by the Norwegian Meteorological Institute (see Tveito et al. 2005).

Site

Altitude m asl

Precipitation Mean annual

Temperature Tetra term

476 436 474 589

2923 2044 1161 600

10.78 10.6 10.5 10.33

780 779 700 815

3029 1848 1356 789

8.67 8.77 9.17 9.14

1133 1213 1097 1208

2725 1925 1321 596

6.58 5.87 6.45 6.17

Lowland LOW4 LOW3 LOW2 LOW1

Intermediate INT4 INT3 INT2 INT1

Alpine ALP4 ALP3 ALP2 ALP1

15

Table 2. Parameter estimates, standard errors and p-values for fixed factors of mixed effects models examining the effects of climate and gap disturbance on pine and spruce seedling emergence, establishment, height and biomass. Temperature is represented by intermediate (= “INT”) and precipitation by category number 2, 3 and 4 from dry to wet. Parameter estimates reflects contrasts with the temperature category “ALP”, the precipitation category “1” and the treatment category “intact vegetation”, which are included in the intercept. * = p < 0.05, ** = p < 0.01, *** = p 0.05 for all interactions with graminoid removal) (not shown).

9

Temperature and precipitation and their interactions were the main determinants of nongraminoid plant species composition (p < 0.001 for all tests). There were no significant interactions between removal and year, or between removal, year and any of the climate variables (p > 0.05 for all tests), meaning that non-graminoid species composition was not affected by the removal treatment (not shown).

10

Discussion

Graminoids, the dominant functional group in semi-natural grasslands, affected grassland community properties by limiting non-graminoid seedling recruitment and plant species richness, while there were no effects on species composition or functional traits. Although competition from graminoids was the main type of interaction with the co-occurring subordinate species, the net outcome of plant-plant interactions also varied with climate, with interacting effects of graminoid removal, temperature and precipitation.

Removal of graminoids increased non-graminoid seedling numbers and species richness across the climate gradients, showing that graminoids in general limit richness and seedling recruitment in the grassland study sites. This is in line with Sasaki and Lauenroth (2011), who found that removal of the dominant grass species Bouteloua gracilis increased species richness in a shortgrass steppe. Competition from established vegetation has also been shown to strongly limit seedling recruitment in perennial grasslands (e.g. Zimmermann et al. 2008 and references therein, Vandvik and Goldberg 2006, Tingstad et al. submitted, Michel et al. in prep.). The negative effect of dominant graminoids on grassland species richness, seemingly mediated through reduced seedling recruitment, is most likely due to competition for light (e.g. Ervin and Wetzel 2002). Light has previously been hypothesized to be a limiting factor for plant growth in the grassland study sites, as plants in intact vegetation seem to stretch for light by allocating resources to height growth (Tingstad et al. submitted, Olsen et al. submitted).

Contrary to our expectations, we did not find a clear switch from facilitation of non-graminoid seedling recruitment or species richness in the cold and dry corner of the climate grid to competition in the warm and wet corner. However, the relative importance of competition did seem to increase with temperature, which is in line with the stress-gradient hypothesis (e.g. Brooker and Callaghan 1998), as interactions with graminoids varied from competition to facilitation along the precipitation gradient in the cold alpine, while competition dominated at all precipitation levels in the warm lowlands. This strong effect of precipitation in the alpine may be due to large amounts of winter precipitation (i.e. snow) in the wet alpine sites, which could have induced the observed shift from competitive to facilitative interactions (Hülber et al. 2011, but see Wipf et al. 2006). Klanderud et al. (submitted) also point to long-lasting

11

snow-cover as an important stress-factor for plants in the cold and wet corner of the climate grid.

The community-level effects of biotic interactions along the climate gradients mainly correspond to previous findings on the population level (Olsen et al. submitted). The relative importance of competition increased with temperature at both the population and community level, showing that it is possible to find consistent patterns across plant organizational levels. However, while competition was the main type of interaction at the community level, the effect of graminoids on the population growth rate of individual species switched from facilitation to competition with increasing temperature (Olsen et al. submitted). Moreover, precipitation seemed to play a more important role for the net outcome of biotic interactions for community properties. Thus, our findings suggest that some, but not all effects of biotic interactions on population-level processes can be scaled up to the community level, and that different factors may have contrasting impacts on different organizational levels. This is only partly in line with Pace (1993), who warns against assuming that patterns and processes at one organizational level can simply be scaled up to the next level (see also Tylianakis et al. 2008, Anthelme et al. 2014, Soliveres and Maestre 2014, Soliveres et al. 2014).

The somewhat different outcome of biotic interactions at different organizational levels may also be due to the short-term nature of our study. While changes in survival and growth of individual plants can be relatively rapid, community-level responses may lag behind. Such a lag in community responses may explain why graminoid removal did not alter non-graminoid functional traits or species composition, as this would require profound changes in the abundance or identity of species making up the plant community. Although three years of graminoid removal affected population growth rates of four study species in the grassland study sites (Olsen et al. submitted), the absolute changes were small, suggesting that alteration of the abundance of well-established perennial species is a slow process. Accordingly, Hollister et al. (2005) found differences in short- and long-term responses of plant communities to climate warming (see also Arft et al. 1999, Walker et al. 2006), indicating that the effect of climate-driven changes in biotic interactions may change in the long term.

In conclusion, our study shows that climate may influence the effect of biotic interactions on seedling recruitment and plant community properties such as species richness. Dominance of competition at high temperatures suggests that competitive interactions with graminoids may 12

further limit non-graminoid species richness and seedling recruitment in semi-natural grasslands as the climate warms, thereby reducing the biodiversity of cold and temperate grassland ecosystems in a future climate. Whereas competitive interactions dominated at high temperatures, results at low temperatures were mixed due to interacting effects of precipitation. While most climate change studies focus solely on temperature, further research is needed to understand the effect of altered precipitation patterns on biotic interactions. This is especially important for alpine habitats, where snow is a main determining factor for plant growth. Finally, our study shows that it is possibly to find a common response to climate warming across organizational levels, but also emphasizes the importance of examining longterm effects on different levels of organization to fully understand how climate change alters biotic interactions between plants.

Acknowledgements

We thank P. Michel, A. Berge, D. Beckmann, B. Bele, S. S. Bruvoll, A. Chételat, A. Coghill, S. Fariñas, A. Halbritter, M. Hamacher, J. Hansen, I. Helvik, M. Jokerud, S. McCall, C. Pötsch, A. V. Skogrand, O. Spearpoint and I. Tween for assistance in the field, S. Fariñas for providing data on total standing biomass, M. R. Boixaderas for trait data collection, O. Skarpaas for helpful comments on the manuscript, and the land owners for granting us access to their grasslands. This study was partially funded by the Norwegian Research Council (NORKLIMA grant #184912/230).

13

References

Abramoff, M. D., Magelhaes, P. J. & Ram, S. J. (2004). Image processing with ImageJ. Biophotonics International, 11: 36-42. Anthelme, F., Cavieres, L. A. & Dangles, O. (2014). Facilitation among plants in alpine environments in the face of climate change. Frontiers in Plant Science. DOI: 10.3389/fpls.2014.00387. Arft, A. M., Walker, M. D., Gurevitch, J., Alatalo, J. M., Bret-Harte, M. S., Dale, M., Diemer, M., Gugerli, F., Henry, G. H. R., Jones, M. H., et al. (1999). Responses of tundra plants to experimental warming: meta-analysis of the International Tundra Experiment. Ecological Monographs, 69: 491-511. Armas, C., Rodríguez-Echeverría, S. & Pugnaire, F. I. (2011). A field test of the stressgradient hypothesis along an aridity gradient. Journal of Vegetation Science, 22: 818827. Bates, D., Maechler, M., Bolker, B. & Walker, S. (2013). lme4: linear mixed-effects models using Eigen and S4. R package version 1.0-4. Belyea, L. R. & Lancaster, J. (1999). Rules within a contingent ecology. Oikos, 86: 402-416. Bertness, M. D. & Callaway, R. (1994). Positive interactions in communities. Trends in Ecology and Evolution, 9: 191-193. Brooker, R. W. & Callaghan, T. V. (1998). The balance between positive and negative plant interactions and its relationship to environmental gradients: a model. Oikos, 81: 196207. Brooker, R. W. (2006). Plant-plant interactions and environmental change. New Phytologist, 171: 271-284. Bruno, J. F., Stachowicz, J. J. & Bertness, M. D. (2003). Inclusion of facilitation into ecological theory. Trends in Ecology and Evolution, 18: 119-125. Butterfield, B. J., Cavieres, L. A., Callaway, R. M., Cook, B. J., Kikvidze, Z., Lortie, C. J., Michalet, R., Pugnaire, F. I., Schöb, C., Xiao, S., et al. (2013). Alpine cushion plants inhibit loss of phylogenetic diversity in severe environments. Ecology Letters, 16: 478-486. Cavieres, L. A., Brooker, R. W., Butterfield, B. J., Cook, B. J., Kikvidze, Z., Lortie, C. J., Michalet, R., Pugnaire, F. I., Schöb, C., Xiao, S., et al. (2014). Facilitative plant interactions and climate simultaneously drive alpine plant diversity. Ecology Letters, 17: 193-202. 14

Chapin III, F. S., Walker, B. H., Hobbs, R. J., Hooper, D. U., Lawton, J. H., Sala, O. E. & Tilman, D. (1997). Biotic control over the functioning of ecosystems. Science, 277: 500-504. Cornelissen, J. H. C., Lavorel, S., Garnier, E., Díaz, S., Buchmann, N., Gurvich, D. E., Reich, P. B., ter Steege, H., Morgan, H. D., van der Heijden, M. G. A., et al. (2003). A handbook of protocols for standardised and easy measurement of plant functional traits worldwide. Australian Journal of Botany, 51: 335-380. Dunnett, N. P. & Grime, J. P. (1999). Competition as an amplifier of short-term vegetation responses to climate: an experimental test. Functional Ecology, 13: 388-395. Elmendorf, S. C., Henry, G. H. R., Hollister, R. D., Björk, R. G., Bjorkman, A. D., Callaghan, T. V., Collier, L. S., Cooper, E. J., Cornelissen, J. H. C., Day, T. A., et al. (2012a). Global assessment of experimental climate warming on tundra vegetation: heterogeneity over space and time. Ecology Letters, 15: 164-175. Elmendorf, S. C., Henry, G. H. R., Hollister, R. D., Björk, R. G., Boulanger-Lapointe, N., Cooper, E. J., Cornelissen, J. H. C., Day, T. A., Dorrepaal, E., Elumeeva, T. G., et al. (2012b). Plot-scale evidence of tundra vegetation change and links to recent summer warming. Nature Climate Change, 2: 453-457. Ervin, G. N. & Wetzel, R. G. (2002). Influence of a dominant macrophyte, Juncus effusus, on wetland plant species richness, diversity, and community composition. Oecologia, 130: 626-636. Fox, J. & Weisberg, S. (2011). An {R} companion to applied regression. 2nd edition. Thousand Oaks: Sage. Fremstad, E. (1997). Vegetasjonstyper i Norge. NINA Temahefte 12. Trondheim: NINA. Goldberg, D. E. & Barton, A. M. (1992). Patterns and consequences of interspecific competition in natural communities: a review of field experiments with plants. The American Naturalist, 139: 771-801. Grime, J. P. (1973). Competitive exclusion in herbaceous vegetation. Nature, 242: 344-347. Grime, J. P. (2001). Plant strategies, vegetation processes, and ecosystem properties. Second edition. Chichester, UK: John Wiley & Sons, Ltd. Gross, N., Kunstler, G., Liancourt, P., de Bello, F., Suding, K. N. & Lavorel, S. (2009). Linking individual response to biotic interactions with community structure: a traitbased framework. Functional Ecology, 23: 1167-1178. Gross, N., Börger, L., Soriano-Morales, S. I., Le Bagousse-Pinguet, Y., Quero, J. L., GarcíaGómez, M., Valencia-Gómez, E. & Maestre, F. T. (2013). Uncovering multiscale 15

effects of aridity and biotic interactions on the functional structure of Mediterranean shrublands. Journal of Ecology, 101: 637-649. Harsch, M. A. & HilleRisLambers, J. (2014). Species distrbutions shift downwards across western North America. Global Change Biology. He, Q., Bertness, M. D. & Altieri, A. H. (2013). Global shifts towards positive species interactions with increasing environmental stress. Ecology Letters, 16: 695-706. HilleRisLambers, J., Adler, P. B., Harpole, W. S., Levine, J. M. & Mayfield, M. M. (2012). Rethinking community assembly through the lens of ecoexistence theory. Annual Review of Ecology, Evolution and Systematics, 43: 227-248. Hollister, R. D., Webber, P. J. & Tweedie, C. E. (2005). The response of Alaskan arctic tundra to experimental warming: differences between short- and long-term responses. Global Change Biology, 11: 525-536. Hülber, K., Bardy, K. & Dullinger, S. (2011). Effects of snowmelt timing and competition on the performance of alpine snowbed plants. Perspectives in Plant Ecology, Evolution and Systematics, 13: 15-26. Kardol, P., Campany, C. E., Souza, L., Norby, R. J., Weltzin, J. F. & Classen, A. T. (2010). Climate change effects on plant biomass alter dominance patterns and community evenness in an experimental old-field system. Global Change Biology, 16: 2676-2687. Kikvidze, Z., Pugnaire, F. I., Brooker, R. W., Choler, P., Lortie, C. J., Michalet, R. & Callaway, R. M. (2005). Linking patterns and processes in alpine plant communities: a global study. Ecology, 86: 1395-1400. Klanderud, K. (2005). Climate change effects on species interactions in an alpine plant community. Journal of Ecology, 93: 127-137. Kleyer, M., Bekker, R. M., Knevel, I. C., Bakker, J. P., Thompson, K., Sonnenschein, M., Poschlod, P., van Groenendael, J. M., Klimes, L., Klimešová, J., et al. (2008). The LEDA Traitbase: a database on life-history traits of the Northwest European flora. Journal of Ecology, 96: 1266-1274. Lepš, J. (1999). Nutrient status, disturbance and competition: an experimental test of relationships in a wet meadow. Journal of Vegetation Science, 10: 219-230. Lepš, J. & Šmilauer, P. (2003). Multivariate analysis of ecological data using CANOCO. Cambridge: Cambridge University Press. Lortie, C. J., Brooker, R. W., Choler, P., Kikvidze, Z., Michalet, R., Pugnaire, F. I. & Callaway, R. M. (2004). Rethinking plant community theory Oikos, 107: 433-438.

16

Luo, Y., Gerten, D., Le Maire, G., Parton, W. J., Weng, E., Zhou, X., Keough, C., Beier, C., Ciais, P., Cramer, W., et al. (2008). Modeled interactive effects of precipitation, temperature, and [CO2] on ecosystem carbon and water dynamics in different climatic zones. Global Change Biology, 14: 1986-1999. Meineri, E., Spindelböck, J. & Vandvik, V. (2013). Seedling emergence responds to both seed source and recruitment site climates: a climate change experiment combining transplant and gradient approaches. Plant Ecology, 214: 607-619. Meineri, E., Skarpaas, O., Spindelböck, J., Bargmann, T. & Vandvik, V. (2014). Direct and size-dependent effects of climate on flowering performance in alpine and lowland herbaceous species. Journal of Vegetation Science, 25: 275-286. Oksanen, J., Blanchet, F. G., Kindt, R., Legendre, P., Minchin, P. R., O'Hara, R. B., Simpson, G. L., Solymos, P., Stevens, M. H. H. & Wagner, H. (2012). vegan: Community Ecology Package. R package version 2.0-5. Olsen, S. L. & Klanderud, K. (2014). Biotic interactions limit species richness in an alpine plant community, especially under experimental warming. Oikos, 123: 71-78. Pace, M. L. (1993). Forecasting ecological responses to global change: the need for largescale comparative studies. In Kareiva, P. M., Kingsolver, J. G. & Huey, R. B. (eds) Biotic interactions and global change. Sunderland: Sinauer Associates Inc. Publishers. R Development Core Team. (2012). R: A language and environment for statistical computing. Vienna: R Foundation for Statistical Computing. Royal Botanic Gardens Kew (2014). Seed Information Database (SID). Version 7.1. To be found on the web page http://data.kew.org/sid/ Sasaki, T. & Lauenroth, W. K. (2011). Dominant species, rather than diversity, regulates temporal stability of plant communities. Oecologia, 166: 761-768. Soliveres, S. & Maestre, F. T. (2014). Plant-plant interactions, environmental gradients and plant diversity: a global synthesis of community-level studies. Perspectives in Plant Ecology, Evolution and Systematics, 16: 154-163. Soliveres, S., Smit, C. & Maestre, F. T. (2014). Moving forward on facilitation research: response to changing environments and effects on the diversity, functioning and evolution of plant communities. Biological Reviews. DOI: 10.1111/brv.12110. Suttle, K. B., Thomsen, M. A. & Power, M. E. (2007). Species interactions reverse grassland responses to changing climate. Science, 315: 640-642. Tilman, D. (1982). Resource competition and community structure. Monographs in Population Biology, vol. 17. Princeton: Princeton University Press. 17

Tilman, D. & Pacala, S. (1993). The maintenance of species richness in plant communities. In Ricklefs, R. E. & Schluter, D. (eds) Species diversity in ecological communities. Chicago: University of Chicago Press. Tilman, D. (1994). Competition and biodiversity in spatially structured habitats. Ecology, 75: 2-16. Tylianakis, J. M., Didham, R. K., Bascompte, J. & Wardle, D. A. (2008). Global change and species interactions in terrestrial ecosystems. Ecology Letters, 11: 1351-1363. Vandvik, V. & Goldberg, D. E. (2006). Sources of diversity in a grassland metacommunity: quantifying the contribution of dispersal to species richness. The American Naturalist, 168: 157-167. Walker, M. D., Wahren, C. H., Hollister, R. D., Henry, G. H. R., Ahlquist, L. E., Alatalo, J. M., Bret-Harte, M. S., Calef, M. P., Callaghan, T. V., Carroll, A. B., et al. (2006). Plant community responses to experimental warming across the tundra biome. Proceedings of the National Academy of Sciences, 103: 1342-1346. Weiher, E., van der Kerf, A., Thompson, K., Roderick, M., Garnier, E. & Eriksson, O. (1999). Challenging Theophrastus: a common core list of plant traits for functional ecology. Journal of Vegetation Science, 10: 609-620. Wipf, S., Rixen, C. & Mulder, C. P. H. (2006). Advanced snowmelt causes shift towards positive neighbour interactions in a subarctic tundra community. Global Change Biology, 12: 1496-1506. Zimmermann, J., Higgins, S. I., Grimm, V., Hoffmann, J., Münkemüller, T. & Linstädter, A. (2008). Recruitment filters in a perennial grassland: the interactive roles of fire, competitors, moisture and seed availability. Journal of Ecology, 96: 1033-1044.

18

Paper IV

Oikos 123: 71–78, 2014 doi: 10.1111/j.1600-0706.2013.00336.x © 2013 The Authors. Oikos © 2013 Nordic Society Oikos Subject Editor: Bente Graae. Accepted 8 May 2013

Biotic interactions limit species richness in an alpine plant community, especially under experimental warming Siri L. Olsen and Kari Klanderud S. L. Olsen ([email protected]) and K. Klanderud, Dept of Ecology and Natural Resource Management, Norwegian Univ. of Life Sciences, PO Box 5003, NO-1432 Ås, Norway.

The determinants of local species richness in plant communities have been the subject of much debate. Is species richness the result of stochastic events such as dispersal processes, or do local environmental filters sort species into communities according to their ecological niches? Recent studies suggest that these two processes simultaneously limit species richness, although their relative importance may vary in space and time. Understanding the limiting factors for species richness is especially important in light of the ongoing global warming, as new species establish in resident plant communities as a result of climate-driven migration. We examined the relative importance of dispersal and environmental filtering during seedling recruitment and plant establishment in an alpine plant community subjected to seed addition and long-term experimental warming. Seed addition increased species richness during the seedling recruitment stage, but this initial increase was cancelled out by a corresponding decrease in species richness during plant establishment, suggesting that environmental filters limit local species richness in the long term. While initial recruitment success of the sown species was related to both abiotic and biotic factors, long-term establishment was controlled mainly by biotic factors, indicating an increase in the relative importance of biotic interactions once plants have germinated in a microhabitat with favourable abiotic conditions. The relative importance of biotic interactions also seemed to increase with experimental warming, suggesting that increased competition within the resident vegetation may decrease community invasibility as the climate warms.

What determines local species richness in plant communities is a central and much debated question in ecology. On one hand, neutral theory states that species richness in a given plant community is mainly governed by stochastic events, such as dispersal from the regional species pool (Caswell 1976, Bell 2001, Hubbell 2001). On the other hand, niche theory claims that diversity is mainly determined by local environmental filters, such as micro-climate and biotic interactions, which sort species into communities according to their ecological niches (Hutchinson 1959, MacArthur and Levins 1967, Chase and Leibold 2003). These theories are not necessarily mutually exclusive, and according to Lortie et al. (2004) local species richness may be simultaneously governed by stochastic processes, abiotic environmental filters such as temperature and soil characteristics, and biotic interactions, which includes both competition and facilitation among plants and interactions with other organisms (see also Belyea and Lancaster 1999). However, the relative importance of the different factors determining species richness may vary in space and time (Lortie et al. 2004). The most common way to determine whether seed dispersal or local environmental filters limit local species richness, is to conduct seed-addition experiments. Meta-analyses

of such seed-addition experiments suggest that seed limitation is widespread both in populations of individual species (Turnbull et al. 2000, Clark et al. 2007) and plant communities as a whole (Myers and Harms 2009), indicating that dispersal from the local and regional species pool is an important limiting factor for species richness. Although the importance of environmental filtering has been hypothesized to increase compared to seed limitation over the lifespan of a species (Clark et al. 2007), the duration of most seed-addition experiments is too short to determine whether the relative importance of seed limitation and environmental filtering change over time (Turnbull et al. 2000, Clark et al. 2007, Myers and Harms 2009). Hence, long-term studies are needed to examine the relative importance of dispersal and environmental filtering for species richness at different stages of the plant life cycle. The relative importance of the different limiting factors for species richness may also change with climate warming, both as a direct response to increased temperatures and indirectly through altering of biotic interactions (Lortie et al. 2004). Such changes may in turn allow new species to establish in resident plant communities, possibly leading to major changes in community species richness and composition. This is already happening in alpine habitats, where 71

species richness is rapidly increasing due to climate-driven upward species migration (Grabherr et al. 1994, Odland et al. 2010, Pauli et al. 2012). Despite the vulnerability of alpine plant communities to the upward range shift of lowland plants, we lack knowledge of the limiting factors for plant establishment in alpine habitats (Eskelinen and Virtanen 2005), especially the relative roles of biotic and abiotic factors (Klanderud and Totland 2007). Short-term seed-addition experiments in alpine plant communities show that seed availability and local environmental conditions simultaneously limit species richness during seedling recruitment (Eskelinen and Virtanen 2005, Klanderud and Totland 2007, Lindgren et al. 2007, Dullinger and Hülber 2011). Moreover, both biotic interactions, such as grazing (Eskelinen and Virtanen 2005, Lindgren et al. 2007) and competition from the resident vegetation (Eskelinen and Virtanen 2005, Lindgren et al. 2007, Dullinger and Hülber 2011), and abiotic factors like temperature (Klanderud and Totland 2007) have been found to affect seedling recruitment. However, little is known about the importance of the different limiting factors for species richness at later life stages and how their relative importance may be modified by climate warming. According to the stress-gradient hypothesis (Brooker and Callaghan 1998), which states that the balance between competition and facilitation may vary along gradients of environmental stress, the role of competition should increase with decreasing severity of the environment, for instance when the microclimate becomes more benign. Here we examine the long-term effects of a seed-addition experiment established in 2000 in an alpine plant community subjected to experimental warming. Klanderud and Totland (2007) found that seed availability limited local species richness during the first years of the study. However, seedling recruitment was also related to different biotic and abiotic factors, suggesting that environmental filtering also play a role in determining species richness in the study system. Further, warming slightly increased seedling establishment, but also tended to increase the role of competition. While Klanderud and Totland (2007) studied the seedling recruitment stage, i.e. the first four years after seed addition (2000–2004), the current study covers later plant life stages, from here on collectively called plant establishment (2004–2011). Our aim is to examine whether seed availability remains the limiting factor for local species richness beyond the seedling stage or if environmental filtering is more important for species richness in the long term, and how the relative importance of seed availability and environmental filters may change with experimental warming. Moreover, to assess how the relative importance of different environmental filters may change with warming, we examine the relationships between long-term species establishment and different biotic and abiotic environmental variables under ambient and elevated temperatures.

Material and methods Study area and experimental design The seed-addition and warming experiment was conducted at Finse (60°36′59″N, 07°31′23″E) in the alpine region of 72

southern Norway between 2000 and 2011. During summer (June–August) Finse has a mean monthly temperature and rainfall of 6.3°C and 89 mm, respectively (Norwegian Meteorological Institute 2011). The study site is located at approximately 1550 m a.s.l. on an exposed ridge of Mt Sanddalsnuten (peak at 1556 m a.s.l.). The bedrock consists mainly of phyllite, supporting a species rich Dryas heath community. The vegetation is dominated by the dwarf shrub Dryas octopetala, and other common species are Bistorta vivipara, Carex rupestris, Carex vaginata, Saussurea alpina and Thalictrum alpinum. The experiment was initiated by Klanderud and Totland (2007) in 2000 to study the relative impacts of seed dispersal and different biotic and abiotic factors on plant species richness. They established 80 main plots, each of which consisted of two 30  60 cm split-plots which were further divided into eighteen 10  10 cm subplots. One of the split-plots in each main plot was randomly selected for seed addition, while the other was used as a control for natural background recruitment. Klanderud and Totland (2007) collected seeds and propagules (from here on collectively called seeds) of 27 species (Table 1) at Mt Sanddalsnuten and randomly varied the number of species added to each split-plot to study the effect of seed species richness on community species richness. Each seed-addition plot received 0, 3, 6, 9, 12, 15, 18, 21, 24 or 27 species as seeds, with eight replicates per level of seedaddition. Eleven of the added species were not present in the plots prior to seed addition (Table 1). To examine whether warming affected the relative importance of dispersal and environmental filtering for species richness, Klanderud and Totland (2007) randomly placed opentop chambers (OTCs) upon half of the main plots. OTCs are hexagonal polycarbonate chambers with an inside diagonal of 1 m used for experimental warming of vegetation (see for instance Marion et al. 1997). A series of experiments in the study area have identified environmental factors which may potentially influence species richness of the Dryas heath. Klanderud (2005) and Klanderud and Totland (2005) found that nutrient addition increased plant growth, and creation of gaps increased seedling recruitment (Klanderud 2010), indicating that abiotic factors such as soil nutrient levels and the cover of bare soil may affect plant establishment and hence species richness. Soil moisture should also influence species establishment in this relatively dry ridge habitat. Further, removal of resident vegetation (Klanderud and Totland 2005) and Dryas (Klanderud 2005) affected plant growth, suggesting that biotic community properties such as the number and cover of vascular plants and cover of the dominant Dryas may influence species richness. In addition, litter may reduce the availability of light and space, thereby decreasing establishment of new species, as shown by for instance Foster and Gross (1998). Prior to seed addition and initiation of the warming treatment, Klanderud and Totland (2007) therefore recorded initial cover of bare soil, litter and Dryas and total vegetation cover in each subplot and calculated average values for all split-plots. They also collected soil samples from each main plot and analyzed for water content, loss on ignition and extractable soil nitrogen (NH4 and NO 3 ). Further, they

Table 1. The 27 species added as seeds, the number of split-plots in which they occurred before seed addition (2000) and the change in plot number between 2000 and 2004 (Klanderud and Totland 2007) and 2004 and 2011 (the current study) under ambient temperature and experimental warming at Finse, Norway. Only seed-addition plots are shown. ∗denotes species not present in the plots prior to seed addition. No. of split-plots 2000 Species Antennaria dioica Anthoxanthum nipponicum∗ Arabis alpina∗ Bistorta vivipara Carex atrofusca Carex capillaris Cerastium alpinum Cerastium cerastoides∗ Epilobium anagallidifolium∗ Erigeron uniflorus Festuca vivipara Leontodon autumnalis∗ Luzula spicata Omalotheca supina∗ Oxyria digyna∗ Oxytropis lapponica Parnassia palustris Poa alpina Potentilla crantzii Ranunculus acris∗ Rumex acetosa∗ Saxifraga cespitosa Saxifraga oppositifolia Silene acaulis Taraxacum sp.∗ Thalictrum alpinum Veronica alpina∗

Δ split-plots 2000–2004

Δ split-plots 2004–2011

warming

ambient

warming

ambient

warming

ambient

18 0 0 40 5 5 12 0 0 13 23 0 11 0 0 7 1 3 23 0 0 1 7 38 0 40 0

19 0 0 39 5 4 22 0 0 19 32 0 6 0 0 18 0 6 25 0 0 2 6 35 0 40 0

6 0 0 0 2 3 11 2 7 17 11 10 9 8 9 11 10 4 9 27 0 0 1 1 21 0 8

0 3 1 1 3 1 6 4 5 13 4 13 16 7 6 3 10 2 1 28 0 1 3 2 19 0 10

6 0 0 0 7 4 13 2 7 8 1 8 7 8 9 12 11 3 2 22 0 1 3 0 19 0 7

1 3 1 2 8 1 15 4 5 18 2 13 5 7 6 10 9 3 3 25 0 1 7 0 17 2 10

recorded the abundance of all vascular plant, lichen and bryophyte species as subplot frequencies per split-plot and assigned all vascular species to functional groups (graminoids, forbs and woody plants). Abundance of all vascular plant species was recorded again in 2004 (Klanderud and Totland 2007) and 2011 (this study). To make the datasets from the different years comparable, we grouped Festuca vivipara and F. ovina, and Antennaria alpina and A. dioica, as non-flowering individuals of these species were hard to distinguish. Taraxacum sp. was not identified to species level. Further details on study design and methods can be found in Klanderud and Totland (2007). Nomenclature follows Lid and Lid (2005). Statistical analyses To examine the effects of seed addition and warming on vascular plant species richness in 2000, 2004 and 2011, we used a linear mixed-effects model. The response variable was total vascular species richness (per split-plot), and fixed factors were seed addition, warming and year, as well as their interactions. To account for the repeated measures and nested design, we estimated a year-specific random intercept for split-plot nested in main plot (Bates 2012). We used likelihood ratio tests to select the final model and determine variable significances. Maximum likelihood was used for the likelihood ratio tests, while restricted maximum

likelihood was used to calculate parameter coefficients of the final model. Where significant interactions were observed, we examined the difference in mean values between the different treatments using paired t-tests. To allow for direct comparisons with the results of Klanderud and Totland (2007), we used multiple linear regressions to examine the relationship between the change in vascular species richness in the seed-addition plots from 2004 to 2011 (i.e. long-term species establishment) and different biotic and abiotic variables. All variables were measured prior to seed addition in 2000, and thus we only test the role of initial environmental conditions for subsequent changes in species richness. The biotic variables were initial vascular species richness, Shannon’s diversity index (including vascular plants, lichens and bryophytes), litter cover (%), total vegetation cover (%) and Dryas cover (%), while the abiotic variables were initial soil moisture (%), soil N content (mg l1), loss on ignition (%) and cover of bare soil (%). We also included the number of species added as seeds, which Klanderud and Totland (2007) found to be a highly significant predictor for species richness following seed addition. Collinearity among environmental variables was assessed by the variance inflation factor (VIF), and loss on ignition was excluded from the analyses as it strongly correlated with soil moisture. We performed separate analyses for the warming and ambient temperature plots, and manual backward elimination was used to subsequently remove non-significant (p  0.05) variables from the model. 73

Extractable soil N content and graminoid species richness were log transformed, cover of bare soil was log(x 1) transformed and litter cover was square-root transformed in all analyses. Preliminary analyses showed that initial vascular species richness was not important for changes in subsequent species richness. However, relationships were found between the change in species richness and initial species richness of different functional groups, and we therefore fitted the final model with initial species richness divided into graminoid, forb and woody (not including Dryas) species richness. All analyses were performed in R ver. 2.13.1 using RStudio ver. 0.94.110. We used the package lme4 (Bates et al. 2011) for the mixed effects models and the car package (Fox and Weisberg 2011) for the VIF diagnostics and partial regression plots.

Results There was no long-term effect of seed addition on species richness. Seed addition increased species richness between 2000 and 2004, but this initial increase was cancelled out by a subsequent decrease in species richness between 2004 and 2011 (Fig. 1). This change in the effect of seed addition over time was reflected in a highly significant interaction between year and seed addition (Table 2), with species richness being significantly higher in the seed-addition plots compared to the control plots in 2004 (t-test, p  0.001) and significantly lower in 2000 (t-test, p  0.01) and 2011 (t-test, p  0.001). The effect of seed addition also differed between temperature treatments (Table 2). Across years seed addition increased species richness under ambient temperature (t-test, p  0.04), but not under experimental warming (t-test, p  0.66), most likely reflecting a stronger increase in species richness following seed addition in the ambient temperature plots (Fig. 1). The relative importance of abiotic and biotic factors measured in the seed-addition plots prior to seed addition seemed to change from the seedling recruitment to the plant establishment stage. While the increase in species richness from 2000 to 2004 was related to both biotic and abiotic factors (Klanderud and Totland 2007; Table 3), the decrease in species richness from 2004 to 2011 was significantly related to biotic factors only (Table 3). Moreover, between

Table 2. Parameter estimates, standard errors (SE) and t-values for fixed factors in the mixed-effects model used to examine the effects of seed addition and warming on vascular plant species richness at Finse, Norway in 2000, 2004 and 2011. Significance levels (∗p  0.05, ∗∗p  0.01, ∗∗∗p  0.001) were assessed by likelihood ratio tests, although not for main effects in the presence of significant interactions. Non-significant (ns) interactions were excluded during model selection and therefore not included in the final model. Fixed effects terms Intercept Year 2004 2011 Seed addition Warming Seed addition  Warming∗ Seed addition  Year∗∗∗ Seed addition  2004 Seed addition  2011 Warming  Year Seed addition  Warming  Year

Coefficient

SE

t

15.56

0.39

39.61

0.80 1.05 0.40 1.04 0.72

0.36 0.39 0.33 0.53 0.35

2.21 2.72 1.21 1.95 2.04

4.85 0.79 ns ns

0.47 0.49

10.39 1.61

2004 and 2011 there were more significant relationships in the experimentally warmed plots than in the ambient temperature plots (Table 3). The explanatory power of the models were equal for the warming and ambient temperature plots (R2  0.54). In the warming plots there was a negative relationship between the change in species richness from 2004 to 2011 and the number of species added as seeds (Fig. 2A), initial Shannon’s diversity index (Fig. 2B), woody species richness (Fig. 2D) and litter cover (Fig. 2E), and a positive relationship with initial graminoid species richness (Fig. 2C) (Table 3). These relationships show that long-term species establishment under elevated temperatures was lower in plots with initially high Shannon’s diversity index, woody species richness and litter cover and higher in plots with initially high graminoid species richness. In the ambient temperature plots there was a negative relationship between the change in species richness and the number of species added as seeds (Fig. 3A) and initial forb species richness (Fig. 3B, Table 3), showing that long-term species establishment was lower in plots with initially high forb species richness. The negative relationship between the change in species richness and the number of species added as seeds found under both temperature regimes shows that longterm species establishment was lower in plots receiving many species as seeds. In the short term (2000–2004) more species recruited in plots where many species had been added as seeds (Klanderud and Totland 2007; Table 3). However, these ‘new’ species were lost in the long term, causing a strong decrease in species richness from 2004 to 2011 in plots which initially received many species as seeds.

Discussion Fig. 1. Vascular plant species richness in 2000, 2004 and 2011 with (triangles) and without (circles) seed addition under ambient temperature (open symbols) and experimental warming (filled symbols) at Finse, Norway. Error bars show mean values  1 SE.

74

Seed addition did not increase species richness of the alpine Dryas heath at Finse in the long term, regardless of temperature regime. Species richness was clearly seed limited during seedling recruitment (2000–2004), as seed addition resulted in an initial increase in species richness (Klanderud

Table 3. Regression coefficients for multiple regressions between the change in vascular plant species richness between 2000 and 2004 (Klanderud and Totland 2007) and 2004 and 2011 (the current study) and different environmental variables in experimentally warmed and ambient temperature plots at Finse, Norway. Only seed-addition plots were included in the analyses, and all environmental variables were measured prior to seed addition. p-values are indicated by asterisks (∗p  0.05, ∗∗p  0.01, ∗∗∗p  0.001). Some variables were not included in the regressions (–) by Klanderud and Totland (2007) due to strong correlations with other variables. In the current study non-significant (ns) variables were excluded during model selection and therefore not included in the final models. 2000–2004 Environmental variables No. of species added Shannon’s diversity index Graminoid species richness Woody species richness Forb species richness Litter cover Total vegetation cover Dryas cover Extractable soil N Soil moisture Bare soil cover

2004–2011

Warming

Ambient

Warming

Ambient

0.51∗∗∗ – 0.54∗∗∗ 0.05 0.31 – – 0.29∗∗ 0.37∗∗ 0.61∗∗∗ 0.20

0.53∗∗∗ – 0.39∗∗

0.27∗∗∗ 12.38∗∗∗ 8.96∗∗ 1.36∗ ns 0.98∗ ns ns ns ns ns

0.30∗∗∗ ns ns ns 0.61∗ ns ns ns ns ns ns

and Totland 2007). However, this initial increase was followed by an equally large decrease in species richness, suggesting that environmental filtering increased in importance over time and was the main limiting factor for species richness during plant establishment (2004–2011). These findings in the alpine align with recent studies from other ecosystems showing reduced relative importance of seed limitation over time (Ehrlén et al. 2006, Paine and Harms 2009, Scott and Morgan 2012; but see Foster and Tilman 2003, Erschbamer et al. 2008). Our results suggest that overcoming the dispersal barrier is not enough for species to establish in the Dryas heath, even in a warmer climate. Instead, environmental filters appear to limit the number of species able to persist in the long term. While both abiotic and biotic filters may limit species richness, our study suggests that their relative importance change over time. Klanderud and Totland (2007) found that abiotic factors seemed to play an important role during seedling recruitment, as the initial increase in species richness was related to abiotic factors such as soil moisture and soil nitrogen content. There were, however, no relationships between the decrease in species richness during long-term species establishment and the measured abiotic factors. Although not explicitly tested, this indicates that while abiotic factors are important for seedling recruitment, biotic factors become relatively more important once plants have germinated in a microhabitat with favourable abiotic conditions. Our findings are in line with recent studies showing increased importance of biotic interactions for growth and survival of adult plants compared to seedlings (Howard and Goldberg 2001, Schiffers and Tielbörger 2006, Leger and Espeland 2010) and indicate that although both abiotic and biotic filters may limit species richness, their relative importance differ between plant life stages. Warming had no direct effect on species richness at alpine Finse in the long term. Klanderud and Totland (2007) noted that the added species established more frequently in the experimentally warmed plots, but we found no direct effects of warming on long-term plant establishment. However, although we did not formally test

0.01 0.31∗∗ – – 0.40∗∗ 0.42∗∗ 0.40∗∗ 0.16

this, warming seemed to indirectly affect species richness in the long term by increasing the relative importance of biotic interactions, as there were more significant relationships between biotic factors and species establishment in the experimentally warmed plots compared to the ambient temperature plots. Most of these relationships were negative, suggesting that the role of competition increased with warming. The sharper increase in species richness following seed addition in the ambient temperature plots also indicates a stronger limitation on seedling recruitment under experimental warming, although increased evaporation inside the OTCs leading to drier surface conditions (Olsen and Klanderud unpubl.) may have influenced seedling survival. An increasing role of competition with increasing temperatures has been shown in other warming experiments (Klein et al. 2004, Klanderud 2005, Klanderud and Totland 2005, 2007), as warming increases growth of individual plants (Arft et al. 1999, Walker et al. 2006, Elmendorf et al. 2012), thereby increasing competition for nutrients and light. Our results thus support the stress-gradient hypothesis, which states that the relative role of competition should increase under more benign environmental conditions (Brooker and Callaghan 1998). One aspect of competition which appeared to be enhanced under elevated temperatures is the negative relationship between diversity and invasibility. In our study initial species diversity seemed to be one of the most important limiting factors for long-term species establishment under experimental warming, as indicated by the strong negative relationship between species establishment and Shannon’s diversity index. This supports the theory that diverse communities are less susceptible to invasion than species poor communities (Elton 1958, Tilman 1997, Levine 2000). Highly diverse communities have few unoccupied niches and thus fewer available resources for new species, and are more likely to contain highly invasion-resistant species (Levine and D’Antonio 1999). An increased role of competition from the resident vegetation in a warmer climate may make alpine plant communities more resistant to invasion by upward-migrating lowland plants. Moreover, 75

Fig. 2. Partial regressions between the change in vascular plant species richness between 2004 and 2011 and the five significant predictor variables – the number of species added as seeds (A), Shannon’s diversity index (B), graminoid species richness (C), woody species richness (D) and litter cover (E) – in experimentally warmed plots with seed addition at Finse, Norway. Plots show the relationship between the change in species richness and the labeled variable taking into account the other variables in the multiple regression model (Table 3), with axes showing the residuals of the labeled variable given the other variables in the model. Graminoid species richness is log transformed and litter cover is square-root transformed. All environmental variables were measured prior to seed addition.

if the negative relationship between invasibility and diversity increases in strength with climate warming, as our results suggest, highly diverse habitats such as the Dryas heath may become increasingly resistant to species invasions. Litter also seemed to reduce invasibility under elevated temperatures, as indicated by a negative relationship between litter cover and long-term species establishment in the experimentally warmed plots. This finding is in line with other studies showing negative effects of litter on species colonization (Tilman 1993, Foster and Gross 1998, Rasran et al. 2007; but see Eskelinen and Virtanen 2005) and 76

species richness (Lord and Lee 2001). Litter reduces light penetration to low-stature plants, acts as a physical barrier to seedlings and shoots, releases allelopathic compounds and facilitates pathogens (Facelli and Pickett 1991). Moreover, litter has been reported to accumulate in response to experimental warming (Klein et al. 2004, Walker et al. 2006, Elmendorf et al. 2012), suggesting that the negative effect of litter on community invasibility may increase with climate warming, further increasing the resistance of alpine plant communities to climate-driven upward range shifts of lowland plants.

Fig. 3. Partial regressions between the change in vascular plant species richness between 2004 and 2011 and the two significant predictor variables – the number of species added as seeds (A) and forb species richness (B) – in ambient temperature plots with seed addition at Finse, Norway. Plots show the relationship between the change in species richness and the labeled variable taking into account the other variable in the multiple regression model (Table 3), with axes showing the residuals of the labeled variable given the other variable in the model. Environmental variables were measured prior to seed addition.

In conclusion, environmental filters played a more important role than dispersal in determining species richness of the Dryas heath community beyond the seedling stage. Our findings emphasize the importance of long-term studies for understanding the processes determining local species richness, as the relative importance of different limiting factors may change between plant life-stages. Moreover, the relative importance of biotic interactions, especially competition, seemed to increase over time and with experimental warming, possibly making the Dryas heath relatively resistant to climate-driven upward range shifts of lowland species. Unlike many other alpine plant communities, the Dryas heath has a dense vegetation cover (∼ 80%) and high species diversity, which promote species interactions and thereby decrease community invasibility. In communities with a lower vegetation cover and less prevalent biotic interactions, dispersal may be the main limiting factor for species richness even in the long term, as found by Erschbamer et al. (2008) in sparsely vegetated alpine glacier forelands. These contrasting results suggest that habitats with less dense vegetation and weaker biotic interactions will be more vulnerable than the Dryas heath to colonization by upward-migrating species in a warmer climate. Acknowledgements – We thank D. Flø for assistance in the field, O. Skarpaas and V. Vandvik for statistical advice and helpful comments on the manuscript and Finse Alpine Research Center for accommodation and hospitality.

References Arft, A. M. et al. 1999. Responses of tundra plants to experimental warming: meta-analysis of the International Tundra Experiment. – Ecol. Monogr. 69: 491–511. Bates, D. 2012. Linear mixed model implementation in lme4. –  cran.r-project.org/web/packages/lme4/vignettes/ Implementation.pdf  Bates, D. et al. 2011. lme4: linear and mixed-effects models using S4 classes. – R package ver. 0.999375-42. Bell, G. 2001. Neutral macroecology. – Science 293: 2413–2418.

Belyea, L. R. and Lancaster, J. 1999. Assembly rules within a contingent ecology. – Oikos 86: 402–416. Brooker, R. W. and Callaghan, T. V. 1998. The balance between positive and negative plant interactions and its relationship to environmental gradients: a model. – Oikos 81: 196–207. Caswell, H. 1976. Community structure: a neutral model analysis. – Ecol. Monogr. 46: 327–354. Chase, J. M. and Leibold, M. A. 2003. Ecological niches: linking classical and contemporary approaches. – Univ. of Chicago Press. Clark, C. J. et al. 2007. Are plant populations seed limited? A critique and meta-analysis of seed addition experiments. – Am. Nat. 170: 128–142. Dullinger, S. and Hülber, K. 2011. Experimental evaluation of seed limitation in alpine snowbed plants. – PLoS ONE 6: e21537. Ehrlén, J. et al. 2006. Long-term assessment of seed limitation in plants: results from an 11-year experiment. – J. Ecol. 94: 1224–1232. Elmendorf, S. C. et al. 2012. Global assessment of experimental climate warming on tundra vegetation: heterogeneity over space and time. – Ecol. Lett. 15: 164–175. Elton, C. S. 1958. The ecology of invasions by animals and plants. – Methuen. Erschbamer, B. et al. 2008. Colonization processes on a central Alpine glacier foreland. – J. Veg. Sci. 19: 855–862. Eskelinen, A. and Virtanen, R. 2005. Local and regional processes in low-productive mountain plant communities: the roles of seed and microsite limitation in relation to grazing. – Oikos 110: 360–368. Facelli, J. M. and Pickett, S. T. A. 1991. Plant litter: its dynamics and effects on plant community structure. – Bot. Rev. 57: 1–32. Foster, B. L. and Gross, K. L. 1998. Species richness in a successional grassland: effects of nitrogen enrichment and plant litter. – Ecology 79: 2593–2602. Foster, B. L. and Tilman, D. 2003. Seed limitation and the regulation of community structure in oak savanna grassland. – J. Ecol. 91: 999–1007. Fox, J. and Weisberg, S. 2011. An {R} companion to applied regression, 2nd edn –  socserv.socsci.mcmaster.ca/jfox/Books/ Companion . Grabherr, G. et al. 1994. Climate effects on mountain plants. – Nature 369: 448.

77

Howard, T. G. and Goldberg, D. E. 2001. Competitive response hierarchies for germination, growth and survival and their influence on abundance. – Ecology 82: 979–990. Hubbell, S. P. 2001. The unified neutral theory of biodiversity and biogeography. – Princeton Univ. Press. Hutchinson, G. E. 1959. Homage to Santa Rosalia, or why are there so many kinds of animals? – Am. Nat. 93: 145–159. Klanderud, K. 2005. Climate change effects on species interactions in an alpine plant community. – J. Ecol. 93: 127–137. Klanderud, K. 2010. Species recruitment in alpine plant communities: the role of species interactions and productivity. – J. Ecol. 98: 1128–1133. Klanderud, K. and Totland, Ø. 2005. The relative importance of neighbours and abiotic environmental conditions for population dynamic parameters of two alpine plant species. – J. Ecol. 93: 493–501. Klanderud, K. and Totland, Ø. 2007. The relative role of dispersal and local interactions for alpine plant community diversity under simulated climate warming. – Oikos 116: 1279–1288. Klein, J. A. et al. 2004. Experimental warming causes large and rapid species loss, dampened by simulated grazing, on the Tibetan Plateau. – Ecol. Lett. 7: 1170–1179. Leger, E. A. and Espeland, E. K. 2010. The shifting balance of facilitation and competition affects the outcome of intra- and interspecific interactions over the life history of California grassland annuals. – Plant Ecol. 208: 333–345. Levine, J. M. 2000. Species diversity and biological invasions: relating local process to community pattern. – Science 288: 852–854. Levine, J. M. and D’Antonio, C. M. 1999. Elton revisited: a review of evidence linking diversity and invasibility. – Oikos 87: 15–26. Lid, J. and Lid, D. T. 2005. Norsk flora, 7th edn. – Det Norske Samlaget. Lindgren, Å. et al. 2007. The impact of disturbance and seed availability on germination of alpine vegetation in the Scandinavian mountains. – Arct. Antarct. Alp. Res. 39: 449–454. Lord, L. A. and Lee, T. D. 2001. Interactions of local and regional processes: species richness in tussock sedge communities. – Ecology 82: 313–318.

78

Lortie, C. J. et al. 2004. Rethinking plant community theory. – Oikos 107: 433–438. MacArthur, R. and Levins, R. 1967. The limiting similarity, convergence and divergence of coexisting species. – Am. Nat. 101: 377–385. Marion, G. M. et al. 1997. Open-top designs for manipulating field temperature in high-latitude ecosystems. – Glob. Change Biol. 3: 20–32. Myers, J. A. and Harms, K. E. 2009. Seed arrival, ecological filters, and plant species richness: a meta-analysis. – Ecol. Lett. 12: 1250–1260. Norwegian Meteorological Institute. 2011. eKlima. Data from weather station 25840 at Finse, normal period 1961–1990. –  www.eklima.met.no  Odland, A. et al. 2010. Increasing vascular plant richness on 13 high mountain summits in southern Norway since the early 1970s. – Arct. Antarct. Alp. Res. 42: 458–470. Paine, C. E. T. and Harms, K. E. 2009. Quantifying the effects of seed arrival and environmental conditions on tropical seedling community structure. – Oecologia 160: 139–150. Pauli, H. et al. 2012. Recent plant diversity changes on Europe’s mountain summits. – Science 336: 353–355. Rasran, L. et al. 2007. Effects of litter removal and mowing on germination and establishment of two fen-grassland species along a productivity gradient. – Folia Geobot. 42: 271–288. Schiffers, K. and Tielbörger, K. 2006. Ontogenetic shifts in interactions among annual plants. – J. Ecol. 96: 336–341. Scott, A. J. and Morgan, J. W. 2012. Dispersal and microsite limitation in Australian old fields. – Oecologia 170: 221–232. Tilman, D. 1993. Species richness of experimental productivity gradients: how important is colonization limitation? – Ecology 74: 2179–2191. Tilman, D. 1997. Community invasibility, recruitment limitation, and grassland biodiversity. – Ecology 78: 81–92. Turnbull, L. A. et al. 2000. Are plant populations seedlimited? A review of seed sowing experiments. – Oikos 88: 225–238. Walker, M. D. et al. 2006. Plant community responses to experimental warming across the tundra biome. – Proc. Natl Acad. Sci. USA 103: 1342–1346.

Paper V

Journal of Ecology 2014, 102, 1129–1137

doi: 10.1111/1365-2745.12292

Exclusion of herbivores slows down recovery after experimental warming and nutrient addition in an alpine plant community Siri L. Olsen* and Kari Klanderud Department of Ecology and Natural Resource Management, Norwegian University of Life Sciences, P.O. Box 5003, N-1432  As, Norway

Summary 1. Global change, such as climate warming and nitrogen deposition, has been predicted to induce non-reversible regime shifts in natural ecosystems. However, we lack knowledge of the potential for recovery from global change perturbations and factors influencing the recovery rate. 2. We examined the recovery of an alpine plant community from a combined warming and nutrient addition experiment, which initially caused profound changes in plant community composition and diversity. We also examined whether the recovery process was affected by herbivory, as the presence of herbivores has been shown to modify the response of plant communities to global change. 3. Many aspects of the alpine plant community had not recovered from experimental warming and nutrient addition 6 years after cessation of the treatments. A persistent shift in vascular plant species composition seemed to inhibit recovery and maintain a community dominated by highly competitive grasses at the expense of the previously dominating dwarf shrub Dryas octopetala, lichens and bryophytes. 4. The exclusion of herbivores decreased the community recovery rate by maintaining unfavourable conditions for lichens and bryophytes. 5. Synthesis. Our findings suggest that a shift in dominance hierarchies in arctic and alpine plant communities due to global change is not readily reversible. Herbivory may, however, facilitate plant community recovery. Key-words: alpine plant community, dominance hierarchies, global change, nutrient addition, plant–climate interactions, plant–herbivore interactions, recovery, regime shift, warming

Introduction Ecosystems may exist in multiple stable states (e.g. Holling 1973; May 1977; Scheffer et al. 2001). The shift between stable states, known as a regime shift, is usually caused by a perturbation which pushes the ecosystem across a threshold or tipping point and into an alternative stable state. This regime shift is often abrupt, where the system suddenly ‘flips’ from one state to another (e.g. Holling 1973). Once a regime shift has taken place, restoring pre-change conditions is not sufficient to make the ecosystem recover (Scheffer et al. 2001; Beisner, Haydon & Cuddington 2003; Bestelmeyer et al. 2011), as the factors initiating the change may be different from the ones maintaining the new stable state (e.g. Rassweiler, Schmitt & Holbrook 2010). Global environmental change has been predicted to induce regime shifts in natural ecosystems (CCSP 2009; Leadley *Correspondence author. E-mail: [email protected]

et al. 2010), causing potentially non-reversible changes. Climate warming and nitrogen deposition are two of the main drivers of global environmental change (e.g. Sala et al. 2000), especially in arctic and alpine habitats, where low temperatures (e.g. Billings & Mooney 1968) and nutrient availability (e.g. Bliss 1971) are important limiting factors for plant growth. Accordingly environmental change experiments in the arctic and alpine have shown profound changes in plant community composition following warming (e.g. Hollister, Webber & Tweedie 2005; Walker et al. 2006; Elmendorf et al. 2012) and nutrient addition (e.g. Chapin et al. 1995; van Wijk et al. 2003; Klanderud & Totland 2005). However, we lack knowledge of whether warming and increased nutrient availability induce a regime shift in arctic and alpine plant communities, or whether the communities can recover if temperature and nutrient availability return to pre-change levels. Studies on ecosystem recovery after experimental nutrient addition have shown that while recovery of soil N cycling

© 2014 The Authors. Journal of Ecology © 2014 British Ecological Society

1130 S. L. Olsen & K. Klanderud and tissue nutrient concentrations is quite rapid (e.g. Boxman, van der Ven & Roelofs 1998; Limpens & Heijmans 2008; O’Sullivan et al. 2011), the recovery of plant and fungal communities is a much slower process (Milchunas & Lauenroth 1995; Quist et al. 1999; Power et al. 2006; Clark & Tilman 2008; Edmondson et al. 2013; Isbell et al. 2013; L.E. Street, N.R. Burns & S.J. Woodin in prep.). Strengbom et al. (2001) found that although species composition of vascular plants in a boreal forest had recovered 47 years after cessation of nutrient addition, there were still effects on fungi and bryophytes, suggesting that the return to the original state is extremely slow. Although arctic and alpine habitats are predicted to be especially sensitive to increased nutrient availability (e.g. Sala et al. 2000), we know of only one study (L.E. Street, N.R. Burns & S.J. Woodin in prep.) examining the recovery of an arctic or alpine plant community after nutrient addition. Moreover, to our knowledge no studies have examined the recovery of plant communities after long-term experimental warming. In arctic and alpine areas experimental warming and nutrient addition has led to a substantial increase in vascular plant biomass, especially of graminoids (e.g. Shaver & Chapin 1986; van Wijk et al. 2003; Klanderud & Totland 2005), which may attract herbivores (Gough, Ramsey & Johnson 2007). Herbivory has been shown to have profound effects on the response of plant communities to global environmental change by inhibiting climate-driven shrub expansion (Olofsson et al. 2009; Ravolainen et al. 2014), biomass increases (Madsen et al. 2011) and upward shifts of plant communities (Speed et al. 2012), and by cancelling out the effects of experimental warming and nutrient addition (e.g. Gough, Ramsey & Johnson 2007; Post & Pedersen 2008; Kaarlej€arvi, Eskelinen & Olofsson 2013). These findings indicate that herbivory may make plant communities remain status quo in the face of environmental change. However, we know of no studies examining whether herbivores influence the rate of recovery of plant communities from environmental change. In this study, we examine the recovery of an alpine plant community following a combined warming and nutrient addition experiment. Klanderud & Totland (2005) and Klanderud (2008) found that experimental nutrient addition, both alone and combined with warming, led to profound changes in alpine plant community species composition and diversity as the vegetation in the study site switched from a species-rich dwarf shrub dominated Dryas octopetala heath to a graminoid-dominated community of low species richness within 4 years. Treatments were ceased after three more years to allow for studies of community recovery. To examine whether grazing had an effect on vegetation recovery, herbivore exclosures were erected upon half of the study plots. The main questions addressed in the current study are (i) has the plant community returned to a pre-change state 6 years after warming and nutrient addition ceased and (ii) is the recovery process affected by herbivory?

Materials and methods STUDY AREA AND EXPERIMENTAL DESIGN

The experiment was conducted at Finse in the alpine region of southwestern Norway (60°360 59″N, 07°310 23″E) between 2000 and 2012. During the summer (June–August) Finse has a mean monthly temperature and rainfall of 6.3 °C and 89 mm respectively (Norwegian Meteorological Institute 2010). The study site is located at approximately 1520 m a.s.l. on a southwest-facing slope of Mt. Sanddalsnuten (peak at 1556 m a.s.l.). The bedrock consists mainly of phyllite, supporting a species rich Dryas heath community. The dwarf shrub Dryas octopetala is the dominating species in the site. Other common vascular plant species are Bartsia alpina, Bistorta vivipara, Carex rupestris, C. vaginata, Festuca vivipara, Potentilla crantzii, Salix herbacea, S. reticulata, Saussurea alpina, Silene acaulis and Thalictrum alpinum, while Cetraria ericetorum, C. islandica, Flavocetraria cuculata and F. nivalis are the most common lichens, and Brachytecium albicans, Campyleum stellatum, Dicranum fuscescens, Distichium capillaceum and Sanionia uncinatus the most common bryophytes. The main mammalian herbivores in the study site are mountain hare (Lepus timidus), Norwegian lemming (Lemmus lemmus) and tundra vole (Microtus oeconomus). Domestic sheep graze in the surrounding areas, and wild reindeer (Rangifer tarandus) are occasionally observed. Grazing pressure is expected to be low in the study site, as it is too exposed and barren to support large populations of herbivores. Moreover, lemming populations at Finse have been low throughout our study period (Kausrud et al. 2008; E. Framstad pers. comm. cited in Barraquand et al. 2014). Although we did not perform a systematic survey of herbivores, we noted few signs of mammalian herbivory damage to plants and low numbers of faecal droppings (S.L. Olsen & K. Klanderud, pers. obs.). To study the impact of warming and increased nutrient availability on alpine vegetation, Klanderud & Totland (2005) randomly established 10 blocks each consisting of 4 100 9 100 cm plots in the study site in 2000. Within each block, experimental warming, nutrient addition, warming combined with nutrient addition and no treatment (control) were randomly assigned to the four plots. Hexagonal polycarbonate open top chambers (OTCs) with an inside diagonal of 1 m were used for the warming treatment (see for instance Marion et al. 1997). The OTCs increased summer air temperature (5 cm above ground) by 1.5 °C and soil temperature (5 cm below ground) by 1.0 °C (Klanderud & Totland 2005). Slow-released NPK fertilizer (10 g N, 2 g P and 8 g K per m2 per growing season) was used to increase nutrient availability. The warming and nutrient addition was carried out for 7 years, starting in 2000. In spring 2007, nutrient addition ceased and the OTCs were taken down to allow for studies of plant community recovery. By then the vegetation subjected to the environmental change treatments (from here on referred to as ‘treatments’), especially nutrient addition and warming combined with nutrient addition, had changed dramatically from a species-rich Dryas heath to a species-poor grass-dominated meadow (Klanderud & Totland 2005; Klanderud 2008). To study the effect of herbivory on community recovery, herbivore exclosures, designed to keep all mammalian herbivores out, were randomly erected upon half of the plots of each treatment in the beginning of the 2007 growing season. The 100 9 100 9 100 cm exclosures were constructed from galvanized net with a mesh size of 12.7 9 12.7 mm and dug into the soil to prevent rodents from entering the plots. The exclosures did not influence snow accumulation or snow cover duration (K. Klanderud, pers. obs.). During the initial experiment, the OTCs were left intact throughout the year and may

© 2014 The Authors. Journal of Ecology © 2014 British Ecological Society, Journal of Ecology, 102, 1129–1137

Plant community recovery from global change 1131 have acted as a barrier to large herbivores (Wookey 2008; Moise & Henry 2010; Kaarlej€arvi, Eskelinen & Olofsson 2013). However, Elmendorf et al. (2012) found no evidence of an OTC exclosure effect in their meta-analysis. Accordingly Post & Pedersen (2008) report that they observed large mammalian herbivores reaching into OTCs on several occasions. Moreover, OTCs do not exclude small mammals such as lemmings and voles (Wookey 2008; Kaarlej€arvi, Eskelinen & Olofsson 2013). Thus, we are confident that the experimental set-up did not affect herbivore access during the initial part of the experiment. Adding nutrients to small plots may have created ‘islands’ of more palatable vegetation which could attract herbivores, thereby artificially increasing the contrast between plots with and without herbivore exclosures (Moise & Henry 2010). However, the nutrient addition plots were not more heavily grazed than others (S.L. Olsen & K. Klanderud, pers. obs.), and we believe that herbivore densities were uniform, although low, throughout the study site. Klanderud & Totland (2005) established a permanent 60 9 60 cm sampling quadrat divided into 36 10 9 10 cm subplots within each plot. The presence of all vascular plant, bryophyte and lichen species were recorded in each subplot in 2000 and 2003 (Klanderud & Totland 2005; Klanderud 2008), 2007 and 2012 (this study). In 2012, we also visually estimated litter cover in each sampling quadrat. As some subplots were damaged, the abundance of each species per plot was calculated by dividing the number of subplots in which the species was present by the total number of intact subplots. Some species were not identified to species level, and some were grouped to make the data sets from the different years comparable (see Table S1 in Supporting Information). Nomenclature follows the Norwegian Species Nomenclature Database (Norwegian Biodiversity Information Centre 2012). Further details on the study design and methods can be found in Klanderud & Totland (2005) and Klanderud (2008).

STATISTICAL ANALYSES

We considered the vegetation to have recovered if plant community properties in the treatment plots were similar to those of the control plots in a given year after cessation of the treatments. We chose to compare the treatment plots to the control plots within each year as preliminary analyses showed that community properties in the control plots had changed during the study period. We used multivariate ordination techniques to examine the trajectory of the species composition of the warming, nutrient addition and warming combined with nutrient addition treatments over time. The full vegetation data matrix (40 permanent plots 9 4 observation time points) was subjected to detrended correspondence analysis (DCA) and global non-metric multidimensional scaling (GNMDS) in parallel, as recommended by Økland (1996). The DCA was run with nonlinear rescaling of axes, detrending by segments and no downweighting of rare species. A two-dimensional GNMDS was run with Bray-Curtis dissimilarity measure, 100 initial configurations, maximum 200 iterations and stress tolerance 107 (sensu Davey et al. 2013), and axes were scaled in half-change units. Solutions were compared using Procrustes permutation test. Correlations between DCA and GNMDS axis 1 and 2 were tested using Kendall’s rank correlation test. Strong correlations between DCA and GNMDS axes (|s| > 0.65) and visual inspection of the ordination plots suggested that similar results were obtained by the two methods. As DCA showed a weak tongue effect, we only present the results of GNMDS. We used constrained ordination to test how the warming, nutrient addition and warming combined with nutrient addition treatments affected species composition throughout the study period. Principal

response curves (PRC) were used to visually assess the relative effect of the treatments over time and the response of the different species. Treatments (levels: warming, nutrient addition and warming combined with nutrient addition) with and without herbivore exclosures and year were used as explanatory variables in the construction of the PRCs. As PRC is a variety of redundancy analysis (RDA) (van den Brink & ter Braak 1999), we used RDA to test for treatment effects on species composition in 2000, 2003, 2007 and 2012 respectively. Treatment and exclosure, as well as their interactions, were used as explanatory variables, while block was used as conditioning variable to control for variation between blocks. Although not erected until 2007, exclosure was included as an explanatory variable in all of the years to test for differences between plots prior to herbivore exclusion. Monte Carlo permutation tests with 999 permutations were used to assess variable significances. Separate models were run for total species composition and species composition of the main functional groups (vascular plants, lichens and bryophytes), as Klanderud & Totland (2005) show that different functional groups responded differently to the initial treatments. The treatment effects on total abundance, species richness and species evenness of the different functional groups were examined using linear mixed-effects models. We used the sum of species abundances for a given functional group as a measure of total functional group abundance (sensu Klanderud & Totland 2005). Treatment (levels: warming, nutrient addition and warming combined with nutrient addition), exclosure and year, as well as their interactions, were used as fixed factors. To account for the nested design and repeated measures, we estimated a random intercept for plot nested in block. We used likelihood ratio tests to select the final models (significance level: P < 0.05) and Markov chain Monte Carlo estimation with 10 000 iterations to assess variable significances. Maximum likelihood (ML) was used for the likelihood ratio tests, while restricted maximum likelihood (REML) was used to calculate parameter coefficients. Initial analyses showed that there were few significant treatment effects on species evenness, and we therefore present results for species abundance and richness only. Total abundance of lichens and bryophytes was square root transformed to attain normality and equal variances. Figures show untransformed data. A similar mixed effects model was used to examine the treatment effects on litter cover. Treatment (levels: warming, nutrient addition and warming combined with nutrient addition) and exclosure were used as fixed factors. Year was not included, as litter cover was measured in 2012 only. The random effects and procedure was identical to the model used for community diversity measures. All analyses were performed in R version 2.15.2 (R Development Core Team 2012) using RStudio version 0.96.331 (RStudio, Inc., Boston, Massachusetts, USA). We used the package lme4 (Bates, Maechler & Bolker 2012) for the mixed effects models, languageR (Baayen 2011) for the Markov chain Monte Carlo estimations and vegan (Oksanen et al. 2012) and MASS (Ripley et al. 2012) for the multivariate analyses.

Results The first GNMDS axis seemed to correlate with nutrient addition, as plots with nutrient addition and warming combined with nutrient addition were displaced primarily along this axis over time, while GNMDS axis 2 seemed to be weakly correlated with warming (Fig. 1). Thus, the change in community species composition over time was mainly driven by the nutrient addition treatment and to some degree by warming.

© 2014 The Authors. Journal of Ecology © 2014 British Ecological Society, Journal of Ecology, 102, 1129–1137

1132 S. L. Olsen & K. Klanderud The RDA showed that total plant community species composition was significantly affected by all treatments in all years after the start of the experiment in 2000, except for warming in 2012, suggesting that the plant community had not recovered 6 years after cessation of the treatments (Table 1 and Fig. 2). Species composition of vascular plants was still affected by all treatments in 2012, although most strongly by nutrient addition and warming combined with nutrient addition (Table 1). There were few persistent effects on vascular plant abundance (Table 2 and Fig. 3a) and species richness (Table 2 and Fig. 3b), suggesting that the lack of recovery of species composition was mainly due to a shift in species abundances, where typical alpine Dryas heath species such as Dryas octopetala and Carex rupestris had been replaced by Festuca grasses and forbs like Cerastium alpinum and Saussurea alpina (Fig. 2). Similarly, lichen species composition was still affected by nutrient addition and warming combined with nutrient addition in 2012 (Table 1). However, this lack of recovery was related to a persistent decrease in abundance of most lichen species (Fig. 2), total lichen abundance (Table 2 and Fig. 3c) and lichen species richness (Table 2 and Fig. 3d). Bryophyte species composition (Table 1 and Fig. 2), abundance (Table 2 and Fig. 3e) and species richness (Table 2 and Fig. 3f) followed similar patterns as for the lichens, but treatment effects were less pronounced. Although plant community composition had not recovered from the treatments by the end of this study, the treatment effects were less strong in 2012 compared to immediately after treatment cessation in 2007, suggesting that the recovery process had started (Table 1). However, the rate of recovery depended on herbivory, as shown by the significant effect of

Fig. 1. Ordination diagram for global nonmetric multidimensional scaling (GNMDS) ordination of mean plant community composition in control, warming, nutrient addition and warming combined with nutrient addition plots during the treatment application (2000–2007) and subsequent recovery (2007–2012) with herbivores present (dashed line) and herbivores excluded (solid line) in an alpine Dryas heath at Finse, Norway. Arrows and symbols show the trajectory of the community under different global change treatments over time, from 2000 to 2003 and 2007 to 2012. Treatments were ceased and herbivore exclosures were erected in spring 2007.

After cessation of the treatments there were clear signs of recovery, as species composition in the treatment plots was displaced towards the control plots in 2012 (Fig. 1). However, the recovery process seemed to be dependent on herbivory, as the displacement towards the control plots was much weaker when herbivores were excluded, especially for the nutrient addition and warming combined with nutrient addition treatments (Fig. 1).

Table 1. F and P-values (significance levels: P < 0.1, *P < 0.05, **P < 0.01, ***P < 0.001) of redundancy analyses (RDA) testing the effects of warming (W), nutrient addition (N), warming combined with nutrient addition (WN) and herbivore exclosures (E) on species composition of the total plant community, vascular plants, lichens and bryophytes at Finse, Norway in 2000, 2003, 2007 and 2012. Treatments were ceased and herbivore exclosures were erected in spring 2007 Total plant community

Vascular plants

Treatment

2000

2003

2007

2012

2000

2003

2007

2012

W N WN E W9E N9E WN 9 E

0.73 0.67 1.09 0.75 0.84 0.68 0.63

2.85* 3.06** 7.88*** 0.82 0.78 0.89 0.99

3.51* 6.41*** 11.99*** 0.54 0.89 1.55 1.26

1.79 4.52*** 5.61*** 2.95** 0.86 1.37 1.24

0.54 0.70 1.02 0.71 0.69 0.84 0.70

1.65 2.80** 3.52** 0.87 0.62 0.82 1.01

3.62** 6.05*** 8.95*** 0.59 0.73 2.00 1.36

2.52* 5.84*** 7.95*** 0.72 0.59 1.41 1.15

Lichens

W N WN E W9E N9E WN 9 E

Bryophytes

2000

2003

2007

2012

2000

2003

2007

2012

0.66 0.44 0.97 1.01 0.68 0.51 0.38

8.20** 5.35** 24.28*** 1.00 1.29 0.58 0.84

4.29* 12.00*** 26.99*** 0.45 1.42 0.39 0.91

1.93 2.11 5.37** 16.87*** 0.79 1.60 2.27

1.04 0.82 1.29 0.58 1.17 0.61 0.74

2.18 1.36 9.61*** 0.34 0.96 1.70 1.05

2.54* 3.26* 10.78*** 0.46 1.02 0.89 1.21

0.48 3.10** 1.62 1.85* 1.35 1.23 1.01

© 2014 The Authors. Journal of Ecology © 2014 British Ecological Society, Journal of Ecology, 102, 1129–1137

Plant community recovery from global change 1133

Fig. 2. Principal response curve (PRC) ordination of mean plant community composition in control, warming, nutrient addition and warming combined with nutrient addition plots during the treatment application (2000–2007) and subsequent recovery (2007–2012) with herbivores present (dashed line) and herbivores excluded (solid line) in an alpine Dryas heath at Finse, Norway. Treatments were ceased and herbivore exclosures were erected in spring 2007 (vertical dotted line). The horizontal grey line represents control plots with herbivores present, to which all other treatments are compared. Note that only PRC axis 1 is shown. The responses of the most common vascular plants (V), lichens (L) and bryophytes (B) are shown to the right. Full species names can be found in Table S1.

herbivore exclosures in 2012 (Table 1). In the presence of herbivores species composition of the treatment plots were displaced towards the control plots, indicating recovery, with the most pronounced effects for the nutrient addition and warming combined with nutrient addition treatments (Fig. 2). There were few signs of recovery in the herbivore exclosures (Fig. 2). The negative effect of herbivore exclusion on community recovery was mainly due to persistent changes in lichen and bryophyte species composition (Table 1), as lichen and bryophyte abundances (Table 2, Figs 3c and 3e) and species richness (Table 2, Figs 3d and 3f) remained low in the herbivore exclosures after cessation of nutrient addition and warming combined with nutrient addition. Further, plant community composition in control plots with exclosures was displaced towards those of the treatment plots (Fig. 2), mainly due to a decrease in lichen (Fig. 3c) and bryophyte (Fig. 3e) abundances, suggesting that herbivory is important for maintaining non-vascular plant communities also under ambient conditions. Litter cover was significantly higher in the combined warming and nutrient addition plots (60.3  22.3%) compared to the control plots (21.9  13.8%) (P < 0.001) in 2012, with a similar trend for nutrient addition (36.4  23.1%) (P < 0.1), but not warming (23.0  8.3%) (P > 0.1). Exclusion of herbivores did not affect litter cover.

Discussion Many aspects of the alpine plant community at Finse had not recovered from experimental warming and nutrient addition 6 years after cessation of the treatments, although our results suggest that the community recovery process had started. This

is in line with previous studies showing that decades may be needed for an ecosystem to return to the original state after a perturbation (Milchunas & Lauenroth 1995; Strengbom et al. 2001; Clark & Tilman 2008; Isbell et al. 2013; L.E. Street, N.R. Burns & S.J. Woodin in prep.). A slow return-time could indicate that the system has been pushed close to a tipping point (Scheffer & Carpenter 2003), and we do not know whether the Dryas heath community will fully recover. We also found that excluding herbivores decreased community recovery rates, suggesting that plant communities with herbivores are more likely to recover from global change perturbations. The lack of recovery was related to persistent changes in composition, abundance and species richness of all functional groups. Vascular plant species composition had not recovered from any of the treatments 6 years after treatment cessation, mainly due to a persistent shift in species abundances. The abundance of some vascular plant species, such as the Festuca grasses, increased strongly in response to nutrient addition and warming combined with nutrient addition, while the previously dominating Dryas decreased, probably as a result of increased competition from fast-growing graminoids and forbs (see Klanderud & Totland 2005). Increased graminoid abundance and biomass at the expense of other functional groups is a common response to warming (e.g. Hollister, Webber & Tweedie 2005; Walker et al. 2006; Elmendorf et al. 2012) and nutrient addition (e.g. Shaver & Chapin 1986; van Wijk et al. 2003; J€agerbrand et al. 2009) in arctic and alpine plant communities. The lack of recovery of vascular species composition after treatment cessation shows that the shift to a more graminoid-dominated community described by Klanderud & Totland (2005) is not readily reversible. Our results are in line with other studies examining plant community recovery from nutrient addition: Quist et al. (1999) found that understorey vascular plant species composition of a boreal forest showed no signs of recovery 8 years after treatment cessation, as the previously dominating Vaccinium dwarf shrubs had been replaced by, amongst others, the graminoid Deschampsia flexuosa (syn. Avenella flexuosa). Similarly, Isbell et al. (2013) found that the graminoid Elymus repens increased strongly in response to nutrient addition and still dominated a previously high-diversity grassland 20 years after termination of the treatment. Together, these findings suggest that a shift in dominance hierarchies within vascular plant communities is a common response that may prevent recovery from increased nutrient availability. Responses of vascular plants to global change treatments may strongly affect non-vascular plants (e.g. Molau & Alatalo 1998), and the persistent change in vascular species composition may thus have inhibited the recovery of lichens and bryophytes in our study site. The slow recovery of the nonvascular plants is in line with previous studies showing that the lichen and bryophyte communities do not readily recover from nutrient addition (Quist et al. 1999; Strengbom et al. 2001; Power et al. 2006; Edmondson et al. 2013). Klanderud & Totland (2005) propose that an increase in vascular plant biomass in the Dryas heath community may have increased

© 2014 The Authors. Journal of Ecology © 2014 British Ecological Society, Journal of Ecology, 102, 1129–1137

1134 S. L. Olsen & K. Klanderud Table 2. Parameter estimates for fixed effects in mixed-effects models examining the effects of warming (W), nutrient addition (N), warming combined with nutrient addition (WN) and herbivore exclosures (E) on abundance and species richness of vascular plants, lichens and bryophytes at Finse, Norway in 2000, 2003, 2007 and 2012. Treatments were ceased and herbivore exclosures were erected in spring 2007. Model selection was performed using likelihood ratio tests, and non-significant interactions were excluded from the final models Vascular plants

Lichens

Bryophytes

Parameters

Abundance

Richness

Abundance

Richness

Abundance

Richness

Intercept 2003 2007 2012 W W:2003 W:2007 W:2012 N N:2003 N:2007 N:2012 WN WN:2003 WN:2007 WN:2012 E E:2003 E:2007 E:2012 W:E W:E:2003 W:E:2007 W:E:2012 N:E N:E:2003 N:E:2007 N:E:2012 WN:E WN:E:2003 WN:E:2007 WN:E:2012

6.98*** 0.29 0.03 0.56 0.71 0.53 1.02* 0.92* 0.32 1.03* 1.14* 1.19** 1.57 0.10 1.43** 0.65 0.84

17.44*** 0.30 1.30 2.10* 1.54 1.40 1.20 0.70 2.84 2.80* 0.50 0.50 2.16 0.50 4.40** 2.10 0.87

1.79*** 0.21 0.34** 0.16 0.20 0.12 0.20 0.27 0.09 0.69*** 0.78*** 0.16 0.08 1.45*** 1.32*** 0.49** 0.12 0.02 0.12 0.22 0.27 0.11 0.05 0.29 0.32 0.15 0.49* 0.70** 0.06 0.30 0.22 0.13

7.59*** 0.80 0.80 1.60 2.82* 0.60 2.40* 3.40** 0.83 3.00* 4.20*** 3.40** 1.79 6.80*** 7.60*** 4.40*** 0.42 0.20 0.40 0.20 3.04 1.00 1.20 1.80 0.05 1.40 1.60 3.00 1.38 1.60 1.60 2.00

1.82*** 0.75*** 0.49*** 0.36* 0.26 0.31 0.37* 0.25 0.23 0.26 0.48** 0.14 0.26 0.90*** 1.41*** 0.44* 0.03 0.04 0.06 0.31*

14.2*** 7.0*** 2.7* 6.3*** 1.6 1.4 3.5* 1.2 0.3 1.0 3.7* 0.8 1.2 5.0** 11.1*** 3.0

2.04

3.89

0.89

3.29

3.14*

6.53*

Significance levels (P < 0.1, *P < 0.05, **P < 0.01, ***P < 0.001) of the final models were assessed by Markov chain Monte Carlo estimation

competition for light, thereby excluding the ground-dwelling lichens and bryophytes, as suggested in many other warming (e.g. Cornelissen et al. 2001; Walker et al. 2006; Lang et al. 2012) and nutrient addition experiments (e.g. Cornelissen et al. 2001; van Wijk et al. 2003; van der Wal, Pearce & Brooker 2005). A lack of recovery of the vascular plant community, with a persistent increase in the abundance of highly competitive graminoids, may therefore have prevented lichen and bryophyte community recovery in our study. Moreover, the increased cover of vascular plant litter in plots with nutrient addition and warming combined with nutrient addition may have buried the lichens and bryophytes, thereby further reducing light availability (van Wijk et al. 2003; Klanderud & Totland 2005). A higher litter cover could also prevent reestablishment of lichens and bryophytes. In addition, lichens and bryophytes are slow-growing compared to vascular plants, suggesting that even if competition and litter cover return to normal levels, recovery may lag behind. The negative effect of persistent changes in the vascular plant community on the recovery of lichens and bryophytes

was sustained by excluding herbivores. Exclusion of herbivores strongly depressed lichen and bryophyte recovery rates, while the presence of herbivores was important for maintaining species composition under ambient conditions, suggesting that lichen and bryophyte communities depend on herbivory. Our results are in line with Virtanen (2000), who found that lichens and bryophytes were absent or heavily reduced in herbivore exclosures, most likely due to competitive exclusion. Herbivores remove plant biomass and may reduce litter accumulation (Virtanen 2000) and increase litter decomposition (Olofsson, Stark & Oksanen 2004), thereby improving conditions for lichens and bryophytes. Herbivore exclosures had no effect on litter cover at Finse, but herbivory may have reduced litter thickness, thereby reducing the shading effect of the litter and promoting re-establishment. Despite the low grazing pressure in the study area and our low sample size, we found clear negative effects of herbivore exclusion on lichen and bryophyte recovery, suggesting that herbivores may have even more pronounced effects in areas with higher herbivore numbers. Too heavy grazing has, however, been

© 2014 The Authors. Journal of Ecology © 2014 British Ecological Society, Journal of Ecology, 102, 1129–1137

Plant community recovery from global change 1135

Fig. 3. Abundance and species richness of vascular plants (a-b), lichens (c-d) and bryophytes (e-f) in control, warming, nutrient addition and warming combined with nutrient addition plots during the treatment application (2000–2007) and subsequent recovery (2007–2012) with herbivores present (dashed line) and herbivores excluded (solid line) in an alpine Dryas heath at Finse, Norway. Treatments were ceased and herbivore exclosures were erected in spring 2007 (vertical dotted line). Error bars show mean values  1 SE. Note differences in y-axis scales.

(a)

(b)

(c)

(d)

(e)

(f)

shown to reduce the cover of lichens (e.g. Eskelinen & Oksanen 2006) and depth of the bryophyte layer (e.g. van der Wal, van Lieshout & Loonen 2001), suggesting that herbivory may negatively affect lichen and bryophyte recovery beyond a certain herbivore pressure threshold. Decomposition of litter resulting from increased vascular plant biomass may also play a key role in the future trajectory of the Dryas heath community, especially following nutrient addition (Klanderud 2008). Rapid decomposition of nitrogenrich litter may sustain high nutrient cycling rates after cessation of nutrient addition (Clark et al. 2009) and thus delay community recovery. However, Soudzilovskaia et al. (2007) found that even though nutrient addition increased litter nutrient concentration in alpine Caucasus, litter decomposition did not increase, mainly due to a strong increase in slowly decomposable graminoid litter. These results correspond well with our findings of graminoid-dominated (S.L. Olsen & K. Klanderud, pers. obs.) litter accumulation in the nutrient addition and warming combined with nutrient addition plots. The immobilization of nutrients in slowly decomposable litter may buffer changes in nutrient cycling rates in the short term (Soudzilovskaia et al. 2007) and thus promote community recovery. However, increased litter cover may simultaneously inhibit recovery by reducing recruitment (Milchunas & Lauenroth

1995; Quist et al. 1999; Isbell et al. 2013; see also Clark & Tilman 2010). In the long-term nutrient cycling rates may increase as nutrient-rich litter is slowly decomposed, thereby maintaining a plant community dominated by vascular plants at the expense of lichens and bryophytes. Nevertheless, decomposition in the alpine is slow, and nutrients may be so gradually released that vascular plants are not favoured, thereby slowly improving conditions for community recovery. As discussed above, these processes may also be affected by herbivory, but also by the soil biota, which changed pronouncedly following the initial treatments (H agvar & Klanderud 2009). Finally, decomposition rates depend on the functional traits of the entire plant community, and all changes in species composition will therefore affect decomposition rates (e.g. Cornelissen et al. 2007; Cornwell et al. 2008), making potential feedbacks via litter extremely difficult to predict. Jones & Schmitz (2009) claim that most ecosystems can recover even from severe perturbations. The alpine Dryas heath community at Finse had not recovered from warming and nutrient addition 6 years after cessation of the treatments, but in this case the recovery rate seemed to depend on herbivory. While there were few signs of recovery when herbivores were excluded, the recovery process had started where herbivores were present, indicating that the plant community may

© 2014 The Authors. Journal of Ecology © 2014 British Ecological Society, Journal of Ecology, 102, 1129–1137

1136 S. L. Olsen & K. Klanderud recover given enough time. The time-scale of the recovery process must be viewed relative to the life span of the species involved (Bestelmeyer et al. 2011), suggesting that a community composed of slow-growing, long-lived alpine species like Dryas may take a long time to return to the original state (but see Olofsson 2006 and Isbell et al. 2013). Thus, we may need to add decades (see Jones & Schmitz 2009) to our relatively short-term experiment before we can draw a definite conclusion about the recovery potential of the Dryas heath. Nonetheless, our study, together with that of Quist et al. (1999) and Isbell et al. (2013) (see also Olofsson 2006), indicates that plant community recovery after a perturbation is slow, perhaps even impossible, once a shift in dominance hierarchies has taken place. In the alpine Dryas heath increased competition and litter cover due to a shift to a more graminoid-dominated community seemed to slow down the recovery of the system after treatment cessation, especially in the herbivore exclosures. Similarly, Isbell et al. (2013) conclude that nutrient addition most likely pushed a high-diversity grassland community into an alternative graminoiddominated low-diversity stable state. With a new dominant species or species group, new mechanisms may determine ecosystem structure and function and thereby prevent or delay a return to the original state, as shown by Rassweiler, Schmitt & Holbrook (2010). However, whether or not a system has been pushed across a threshold and into a new stable state is hard to judge. Ecologists and philosophers of science have not agreed on how to identify tipping points, which time-scale to consider, or how different a state must be to be deemed truly alternative (Beisner, Haydon & Cuddington 2003; Jones & Schmitz 2009). Even after 20 years with few signs of recovery, Isbell et al. (2013) could not rule out the possibility that their system could be slowly returning to the original state. In practical ecology, tipping points seem difficult or impossible to recognize, except from in hindsight. In conclusion, our study shows that recovery of arctic and alpine plant communities from global change is at best a long-term process, especially where grazing pressure is low. While nitrogen deposition has the potential to decrease over time and thereby allow for plant community recovery, climate warming is not readily reversible. Although effects of experimental warming was less strong and more transient than the effects of nutrient addition in our experiment, global warming will continue to increase in the coming decades, with steadily increasing effects on arctic and alpine plant communities.

Acknowledgements We thank M. Jokerud for conducting the bryophyte survey, O. Skarpaas, V. Vandvik, the editor, Nigel G. Yoccoz, and anonymous reviewers for helpful comments on the manuscript, and Finse Alpine Research Center for accommodation and hospitality.

Data accessibility All data are available at the figshare online data repository: http://dx.doi.org/10. 6084/m9.figshare.1066696 (Olsen & Klanderud 2014).

References Baayen, R.H. (2011) languageR: data sets and functions with “Analyzing linguistic data: a practial introduction to statistics”. R package version 1.4. Barraquand, F., Høye, T.T., Henden, J.-A., Yoccoz, N.G., Gilg, O., Schmidt, N.M., Sittler, B. & Ims, R.A. (2014) Demographic response of a site-faithful territorial predator to its fluctuating prey: long-tailed skuas and arctic lemmings. Journal of Animal Ecology, 83, 375–387. Bates, D., Maechler, M. & Bolker, B. (2012) lme4: linear and mixed-effects models using S4 classes. R package version 0.999999-0. Beisner, B.E., Haydon, D.T. & Cuddington, K. (2003) Alternative stable states in ecology. Frontiers in Ecology, 1, 376–382. Bestelmeyer, B.T., Ellison, A.M., Fraser, W.R., Gorman, K.B., Holbrook, S.J., Laney, C.M., Ohman, M.D., Peters, D.P.C., Pillsbury, F.C., Rassweiler, A., Schmitt, R.J. & Sharma, S. (2011) Analysis of abrupt transitions in ecological systems. Ecosphere, 2, 1–26. Billings, W.D. & Mooney, H.A. (1968) The ecology of arctic and alpine plants. Biological Reviews, 43, 481–529. Bliss, L.C. (1971) Arctic and alpine plant life cycles. Annual Review of Ecology and Systematics, 2, 405–438. Boxman, A.W., van der Ven, P.J.M. & Roelofs, J.G.M. (1998) Ecosystem recovery after a decrease in nitrogen input to a Scots pine stand at Ysselsteyn, the Netherlands. Forest Ecology and Management, 101, 155–163. van den Brink, P.J. & ter Braak, C.J.F. (1999) Principal response curves: analysis of time-dependent multivariate responses of biological community to stress. Environmental Toxicology and Chemistry, 18, 128–148. CCSP (2009) Thresholds of Climate Change in Ecosystems. A Report by the U.S. Climate Change Science Program and the Subcommittee on Global Change Research. U.S. Geological Survey, Reston. Chapin, F.S. III, Shaver, G.R., Giblin, A.E., Nadelhoffer, K.J. & Laundre, J.A. (1995) Responses of Arctic tundra to experimental and observed changes in climate. Ecology, 76, 694–711. Clark, C.M. & Tilman, D. (2008) Loss of plant species after chronic low-level nitrogen deposition to prairie grasslands. Nature, 451, 712–715. Clark, C.M. & Tilman, D. (2010) Recovery of plant diversity following N cessation: effects of recruitment, litter, and elevated N cycling. Ecology, 91, 3620–3630. Clark, C.M., Hobbie, S.E., Venterea, R. & Tilman, D. (2009) Long-lasting effects on nitrogen cycling 12 years after treatments cease despite minimal long-term nitrogen retention. Global Change Biology, 15, 1755–1766. Cornelissen, J.H.C., Callaghan, T.V., Alatalo, J.M., Michelsen, A., Graglia, E., Hartley, A.E. et al. (2001) Global change and arctic ecosystems: is lichen decline a function of increases in vascular plant biomass? Journal of Ecology, 89, 984–994. Cornelissen, J.H.C., van Bodegom, P.M., Aerts, R., Callaghan, T.V., van Logtestijn, R.S.P., Alatalo, J. et al. (2007) Global negative vegetation feedback to climate warming responses of leaf litter decomposition rates in cold biomes. Ecology Letters, 10, 619–627. Cornwell, W.K., Cornelissen, J.H.C., Amatangelo, K., Dorrepaal, E., Eviner, V.T., Godoy, O. et al. (2008) Plant species traits are the predominant control on litter decomposition rates within biomes worldwide. Ecology Letters, 11, 1065–1071. Davey, M.L., Heegaard, E., Halvorsen, R., Kauserud, H. & Ohlson, M. (2013) Amplicon-pyrosequencing-based detection of compositional shifts in bryophyte-associated fungal communities along an elevation gradient. Molecular Ecology, 22, 368–383. Edmondson, J., Terribile, E., Carroll, J.A., Price, E.A.C. & Caporn, S.J.M. (2013) The legacy of nitrogen pollution in heather moorlands: ecosystem response to simulated decline in nitrogen deposition over seven years. Science of the Total Environment, 444, 138–144. Elmendorf, S.C., Henry, G.H.R., Hollister, R.D., Bj€ ork, R.G., Bjorkman, A.D., Callaghan, T.V. et al. (2012) Global assessment of experimental climate warming on tundra vegetation: heterogeneity over space and time. Ecology Letters, 15, 164–175. Eskelinen, A. & Oksanen, J. (2006) Changes in the abundance, composition and species richness of mountain vegetation in relation to summer grazing by reindeer. Journal of Vegetation Science, 17, 245–254. Gough, L., Ramsey, E.A. & Johnson, D.R. (2007) Plant-herbivore interactions in Alaskan arctic tundra change with soil nutrient availability. Oikos, 116, 407–418. H agvar, S. & Klanderud, K. (2009) Effects of simulated environmental change on alpine soil arthropods. Global Change Biology, 15, 2972–2980. Holling, C.S. (1973) Resilience and stability of ecological systems. Annual Review of Ecology and Systematics, 4, 1–23.

© 2014 The Authors. Journal of Ecology © 2014 British Ecological Society, Journal of Ecology, 102, 1129–1137

Plant community recovery from global change 1137 Hollister, R.D., Webber, P.J. & Tweedie, C.E. (2005) The response of Alaskan arctic tundra to experimental warming: differences between short- and longterm responses. Global Change Biology, 11, 525–536. Isbell, F., Tilman, D., Polasky, S., Binder, S. & Hawthorne, P. (2013) Low biodiversity state persists two decades after cessation of nutrient enrichment. Ecology Letters, 16, 454–460. J€agerbrand, A.K., Alatalo, J.M., Chrimes, D. & Molau, U. (2009) Plant community responses to 5 years of simulated climate change in meadow and heath ecosystems at a subarctic-alpine site. Oecologia, 161, 601–610. Jones, H.P. & Schmitz, O.J. (2009) Rapid recovery of damaged ecosystems. PLoS ONE, 4, e5653. Kaarlej€arvi, E., Eskelinen, A. & Olofsson, J. (2013) Herbivory prevents positive responses of lowland plants to warmer and more fertile conditions at high altitudes. Functional Ecology, 27, 1244–1253. Kausrud, K.L., Mysterud, A., Steen, H., Vik, J.O., Østbye, E., Cazelles, B., Framstad, E., Eikeset, A.M., Mysterud, I., Solhøy, T. & Stenseth, N.C. (2008) Linking climate change to lemming cycles. Nature, 456, 93–97. Klanderud, K. (2008) Species-specific responses of an alpine plant community under simulated environmental change. Journal of Vegetation Science, 19, 363–372. Klanderud, K. & Totland, Ø. (2005) Simulated climate change altered dominance hierarchies and diversity of an alpine biodiversity hotspot. Ecology, 86, 2047–2054. Lang, S.I., Cornelissen, J.H.C., Shaver, G.R., Ahrens, M., Callaghan, T.V., Molau, U., ter Braak, C.J.F., H€olzer, A. & Aerts, R. (2012) Arctic warming on two continents has consistent negative effects on lichen diversity and mixed effects on bryophyte diversity. Global Change Biology, 18, 1096– 1107. Leadley, P., Pereira, H.M., Alkemade, R., Fernandez-Manjarres, J.F., Proenҫa, V., Scharlemann, J.P.W. & Walpole, M.J. (2010) Biodiversity scenarios: projections of the 21st century change in biodiversity and associated ecosystem services. Secretariat of the Convention on Biological Diversity, Montreal. Limpens, J. & Heijmans, M.M.P.D. (2008) Swift recovery of Sphagnum nutrient concentrations after excess supply. Oecologia, 157, 153–161. Madsen, J., Jaspers, C., Tamstorf, M., Mortensen, C.E. & Riget, F. (2011) Long-term effects of grazing and global warming on the composition and carrying capacity of graminoid marshes for moulting geese in East Greenland. Ambio, 40, 638–649. Marion, G.M., Henry, G.H.R., Freckman, D.W., Johnstone, J., Jones, G., Jones, M.H., Levesque, E., Molau, U., Mølgaard, P., Parsons, A.N., Svoboda, J. & Virginia, R.A. (1997) Open-top designs for manipulating field temperature in high-latitude ecosystems. Global Change Biology, 3, 20–32. May, R.M. (1977) Thresholds and breakpoints in ecosystems with a multiplicity of stable states. Nature, 269, 471–477. Milchunas, D.G. & Lauenroth, W.K. (1995) Inertia in plant community structure: state changes after cessation of nutrient-enrichment stress. Ecological Applications, 5, 452–458. Moise, E.R.D. & Henry, H.A.L. (2010) Like moths to a street lamp: exaggerated animal densities in plot-level global change field experiments. Oikos, 119, 791–795. Molau, U. & Alatalo, J. (1998) Responses of subarctic-alpine plant communities to simulated environmental change: biodiversity of bryophytes, lichens and vascular plants. Ambio, 27, 322–329. Norwegian Biodiversity Information Centre (2012) Norwegian Species Names Database. Available at www2.artsdatabanken.no/artsnavn/Contentpages/ Hjem.aspx. Norwegian Meteorological Institute (2010) eKlima. Data from weather station 25840 at Finse, normal period 1961-1990. Available at eklima.met.no. Økland, R.H. (1996) Are ordination and constrained ordination alternative or complementary strategies in general ecological studies? Journal of Vegetation Science, 7, 289–292. Oksanen, J., Blanchet, F.G., Kindt, R., Legendre, P., Minchin, P.R., O’Hara, R.B., Simpson, G.L., Solymos, P., Stevens, M.H.H. & Wagner, H. (2012) vegan: Community Ecology Package. R package version 2.0-5. Olofsson, J. (2006) Short- and long-term effects of changes in reindeer grazing pressure on tundra heath vegetation. Journal of Ecology, 94, 431–440. Olofsson, J., Stark, S. & Oksanen, L. (2004) Reindeer influence on ecosystem processes in the tundra. Oikos, 105, 386–396. Olofsson, J., Oksanen, L., Callaghan, T., Hulme, P.E., Oksanen, T. & Suominen, O. (2009) Herbivores inhibit climate-driven shrub expansion on the tundra. Global Change Biology, 15, 2681–2693. Olsen, S.L. & Klanderud, K. (2014) Data for the Journal of Ecology paper ‘Exclusion of herbivores slows down recovery after experimental warming and nutrient addition in an alpine plant community’. figshare, http://dx.doi. org/10.6084/m9.figshare.1066696.

O’Sullivan, O.S., Horswill, P., Phoenix, G.K., Lee, J.A. & Leake, J.R. (2011) Recovery of soil nitrogen pools in species-rich grasslands after 12 years of simulated pollutant nitrogen deposition: a 6-year experimental analysis. Global Change Biology, 17, 2615–2628. Post, E. & Pedersen, C. (2008) Opposing plant community responses to warming with and without herbivores. Proceedings of the National Academy of Sciences, 105, 12353–12358. Power, S.A., Green, E.R., Barker, C.G., Bell, J.N.B. & Ashmore, M.R. (2006) Ecosystem recovery: heathland response to a reduction in nitrogen deposition. Global Change Biology, 12, 1241–1252. Quist, M.E., N€asholm, T., Lindeberg, J., Johannisson, C., H€ ogbom, L. & H€ ogberg, P. (1999) Responses of a nitrogen-saturated forest to a sharp decrease in nitrogen input. Journal of Environmental Quality, 28, 1970–1977. R Development Core Team (2012) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna. Rassweiler, A., Schmitt, R.J. & Holbrook, S.J. (2010) Triggers and maintenance of multiple shifts in the state of a natural community. Oecologia, 164, 489–498. Ravolainen, V.T., Br athen, K.A., Yoccoz, N.G., Nguyen, J.K. & Ims, R.A. (2014) Complementary impacts of small rodents and semi-domesticated ungulates limit tall shrub expansion in the tundra. Journal of Applied Ecology, 51, 234–241. Ripley, B., Venables, B., Hornik, K., Gebhardt, A. & Firth, D. (2012) MASS: support functions and datasets for Venables and Ripley’s MASS. R package version 7.3-22. Sala, O.E., Chapin, F.S. III, Armesto, J.J., Berlow, E., Bloomfield, J., Dirzo, R. et al. (2000) Global biodiversity scenarios for the year 2100. Science, 287, 1770–1774. Scheffer, M. & Carpenter, S.R. (2003) Catastrophic regime shifts in ecosystems: linking theory to observation. Trends in Ecology and Evolution, 18, 648–656. Scheffer, M., Carpenter, S., Foley, J.A., Folke, C. & Walker, B. (2001) Catastrophic shifts in ecosystems. Nature, 413, 591–596. Shaver, G.R. & Chapin, F.S. III (1986) Effect of fertilizer on production and biomass of tussock tundra, Alaska, U.S.A. Arctic and Alpine Research, 18, 261–268. Soudzilovskaia, N.A., Onipchenko, V.G., Cornelissen, J.H.C. & Aerts, R. (2007) Effects of fertilization and irrigation on ‘foliar afterlife’ in alpine tundra. Journal of Vegetation Science, 18, 755–766. Speed, J.D.M., Austrheim, G., Hester, A.J. & Mysterud, A. (2012) Elevational advance of alpine plant communities is buffered by herbivory. Journal of Vegetation Science, 23, 617–625. Strengbom, J., Nordin, A., N€asholm, T. & Ericson, L. (2001) Slow recovery of boreal forest ecosystem following decreased nitrogen input. Functional Ecology, 15, 451–457. Virtanen, R. (2000) Effects of grazing on above-ground biomass on a mountain snowbed, NW Finland. Oikos, 90, 295–300. van der Wal, R., Pearce, I.S.K. & Brooker, R.W. (2005) Mosses and the struggle for light in a nitrogen-polluted world. Oecologia, 142, 159–168. van der Wal, R., van Lieshout, S.M.J. & Loonen, M.J.J.E. (2001) Herbivore impact on moss depth, soil temperature and arctic plant growth. Polar Biology, 24, 29–32. Walker, M.D., Wahren, C.H., Hollister, R.D., Henry, G.H.R., Ahlquist, L.E., Alatalo, J.M. et al. (2006) Plant community responses to experimental warming across the tundra biome. Proceedings of the National Academy of Sciences, 103, 1342–1346. van Wijk, M.T., Clemmensen, K.E., Shaver, G.R., Williams, M., Callaghan, T.V., Chapin, F.S. III et al. (2003) Long-term ecosystem level experiments at Toolik Lake, Alaska, and at Abisko, Northern Sweden: generalizations and differences in ecosystem and plant type responses to global change. Global Change Biology, 10, 105–123. Wookey, P.A. (2008) Experimental approaches to predicting the future of tundra plant communities. Plant Ecology & Diversity, 1, 299–307. Received 5 February 2014; accepted 24 June 2014 Handling Editor: Will Cornwell

Supporting Information Additional Supporting Information may be found in the online version of this article: Table S1. Abbreviations, full species names and functional groups of all species recorded in the experimental plots.

© 2014 The Authors. Journal of Ecology © 2014 British Ecological Society, Journal of Ecology, 102, 1129–1137

Abbreviation Agr.mer Ant.dio Ant.nip Ara.alp Ast.alp Bar.alp Bis.viv Bot.lun Cam.rot Car.atr Car.atro Car.cap Car.rup Car.sp. Cer.alp Com.ten Dra.nor Dry.oct Emp.nig Equ.var Eri.uni Eup.wet Fes.spp. Gen.niv Har.hyp Hie.alp

Full species name Agrostis mertensii Antennaria dioica Anthoxanthum nipponicum Arabis alpina Astragalus alpinus Bartsia alpina Bistorta vivipara Botrychium lunaria Campanula rotundifolia Carex atrata Carex atrofusca Carex capillaris Carex rupestris Carex spp. Cerastium alpinum Comastoma tenellum Draba norvegica Dryas octopetala Empetrum nigrum Equisetum variegatum Erigeron uniflorus Euphrasia wettsteinii Festuca spp. Gentiana nivalis Harrimanella hypnoides Hieracium alpinum

Functional group Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant F. ovina, F. rubra and F. vivipara

C. bigelowii and C. vaginata

Comment

comment). Nomenclature follows the Norwegian Species Nomenclature Database (Norwegian Biodiversity Information Centre 2012).

Dryas heath at Finse, Norway, between 2000 and 2012. Some species were not identified to species level, and some were grouped (see

Table S1. Abbreviations, full species names and functional groups of the species found in plots with simulated environmental change in an alpine

Abbreviation Hup.sel Jun.big Jun.tri Leo.aut Luz.spp. Min.bif Oma.sup Oxy.dig Oxy.lap Par.pal Phl.alp Poa.alp Pot.cra Pyr.min Ran.acr Rum.ace Sal.her Sal.ret Sau.alp Sax.aiz Sax.cer Sax.ces Sax.opp Sel.sel Sil.aca Sil.wah Tar.sp Tha.alp Tof.pus Tri.spi Vac.uli

Full species name Huperzia selago Juncus biglumis Juncus trifidus Leontodon autumnalis Luzula spp. Minuartia biflora Omalotheca supina Oxyria digyna Oxytropis lapponica Parnassia palustris Phleum alpinum Poa alpina Potentilla crantzii Pyrola minor Ranunculus acris Rumex acetosa Salix herbacea Salix reticulata Saussurea alpina Saxifraga aizoides Saxifraga cernua Saxifraga cespitosa Saxifraga oppositifolia Selaginella selaginoides Silene acaulis Silene wahlbergella Taraxacum sp. Thalictrum alpinum Tofieldia pusilla Trisetum spicatum Vaccinium uliginosum

Functional group Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant Vascular plant L. multiflora and L. spicata

Comment

Abbreviation Ver.alp Vis.alp Ale.och Bry.div Cet.del Cet.spp. Cla.arb Cla.ran Cla.spp. Cla.unc Fla.cuc Fla.niv Och.fri Pel.aph Pel.spp. Per.pan Rin.tur Sol.sac Sph.glo Ste.spp. Tha.ver Vul.jun Bar.spp. Bar sp. Ble.tri Bra.alb Bry.spp. Cam.chr Cam.ste Cep/Lop.spp. Cyn.sp.

Full species name Veronica alpina Viscaria alpina Alectoria ochroleuca Bryocaulon divergens Cetrariella delisei Cetraria spp. Cladonia arbuscula Cladonia rangiferina Cladonia spp. Cladonia uncialis Flavocetraria cucullata Flavocetraria nivalis Ochrolechia frigida Peltigera aphthosa Peltigera spp. Pertusaria panyrga Rinodina turfacea Solorina saccata Sphaerophorus globosus Stereocaulon spp. Thamnolia vermicularis Vulpicida juniperinus Barbilophozia spp. Barbula sp. Blepharostoma trichophyllum Brachythecium albicans Bryum spp. Campyliadelphus chrysophyllus Campylium stellatum Cephalozia and Lophozia spp. Cynodontium sp.

Functional group Vascular plant Vascular plant Lichen Lichen Lichen Lichen Lichen Lichen Lichen Lichen Lichen Lichen Lichen Lichen Lichen Lichen Lichen Lichen Lichen Lichen Lichen Lichen Bryophyte Bryophyte Bryophyte Bryophyte Bryophyte Bryophyte Bryophyte Bryophyte Bryophyte Peltigera species except P. aphtosa

Cladonia subg. cladonia species except C. uncialis

C. ericetorum and C. islandica

Comment

Abbreviation Dic.spp. Dic-lla.spp. Dis.cap Dit.fle Enc.sp. Fis.spp. Hyl.spl/Ple.sch Hyp.spp. Jun.spp. Mar.pol. Mee.uli Myu.ten Onc.wah Ort.int Pla.por Poh.spp. Pol.spp. Pre.qua Pti.cil Rac.spp. Rhy.rug Sae.gla San.unc Sca.spp. Syn.spp. Tet.mni Tor.tor

Full species name Dicranum spp. Dicranella spp. Distichium capillaceum Ditrichum flexicaule Encalypta sp. Fissidens spp. Hylocomium splendens and Pleurozium schreberi Hypnum spp. Jungermannia spp. Marchantia polymorpha Meesia uliginosa Myurella tenerrima Oncophorus wahlenbergii Orthothecium intricatum Plagiochila porelloides Pohlia spp. Polytrichum spp. Preissia quadrata Ptilidium ciliare Racomitrium spp. Rhytidium rugosum Saelania glaucescens Sanionia uncinata Scapania spp. Syntrichia spp. Tetraplodon mnioides Tortella tortuosa

Functional group Bryophyte Bryophyte Bryophyte Bryophyte Bryophyte Bryophyte Bryophyte Bryophyte Bryophyte Bryophyte Bryophyte Bryophyte Bryophyte Bryophyte Bryophyte Bryophyte Bryophyte Bryophyte Bryophyte Bryophyte Bryophyte Bryophyte Bryophyte Bryophyte Bryophyte Bryophyte Bryophyte

Comment