Cover photo: A mix of charophytes, flooded terrestrial vascular plants and the amphibious species Littorella uniflora all of which are treated in the present thesis. Photo: Ole Pedersen

F R E S H W A T E R B I O L O G I C A L L A B O R A T O R Y F A C U L T Y O F S C I E N C E

U N I V E R S I T Y O F C O P E N H A G E N

Distribution and functional traits of charophytes and vascular plants

- Biodiversity in aquatic systems and along flooding gradients

PhD thesis by

Lars Båstrup-Spohr

Academic advisor: Professor Kaj Sand Jensen

Submitted: 07/06/13

This thesis has been submitted to the PhD School of The Faculaty of Science, University of Copenhagen

1

  

Author        Lars Båstrup‐Spohr 

        Freshwater Biological Laboratory 

University of Copenhagen 

Denmark 

lbaastrupspohr@bio.ku.dk 

 

Supervisor      Kaj Sand‐Jensen, Professor 

        Freshwater Biological Laboratory 

University of Copenhagen 

Denmark 

 

Committee      Dean Jacobsen, Associate Professor (Chair) 

        Freshwater Biological Laboratory 

University of Copenhagen 

Denmark 

 

Henrik Balslev, Professor 

Department of Bioscience ‐ Ecoinformatics and Biodiversity 

Aarhus University 

Denmark 

 

Esperanҫa Gacia, Senior Scientist, Ph.D.  Biogeodynamics and Biodiversity Group 

Centre d’Estudis Avançats de Blanes, Consejo Superior de 

Investigaciones Cientificas (CEAB‐CSIC) 

Spain 

 

Financial support    The Faculty of Science 

University of Copenhagen 

Center for Lake Restoration (CLEAR) A Villum Kann Rasmussen 

Center of Excellence  

 

 

 

 

 

 

   

2

List of papers  

  

List of papers 

 

The thesis includes the following papers: 

 

I. Baastrup‐Spohr, L., Iversen, L. L., Dahl‐Nielsen, J. & Sand‐Jensen, K (2013) Seventy years 

of changes in the abundance of Danish charophytes. Freshwater Biology in press 

II. Baastrup‐Spohr, L., Iversen, L. L., Borum, J. & Sand‐Jensen, K (2013) Niche specialization 

and functional traits regulate abundance of Scandinavian charophytes. Submitted to 

Freshwater Biology 

III. Baastrup‐Spohr, L., Sand‐Jensen, K., Nicolajsen S. V., & Bruun H. H. (2013) From pond to 

alvar: predicting plant community composition along a steep hydrological gradient. 

In revision in Journal of Ecology 

IV. Baastrup‐Spohr, L., Møller, C. L. & Sand‐Jensen, K. (2013) Water level fluctuations affect 

plant growth and carbon dynamics of isoetid populations. Manuscript 

Appendix I  Baastrup‐Spohr, L., Dahl‐Nielsen, J. & Sand‐Jensen, K. (2013) Kransnålalger rummer 

mange truede arter (in Danish). Urt 37(2); 66‐70 

Appendix II  Sand‐Jensen, K., Baastrup‐Spohr, L., Winkel, A., Møller, C. L., Borum, J., Brodersen, 

K. P., Lindell, T. & Stæhr, P. A. (2010) Ett kalkbrott på Ölands alvar – en stenöken 

med knivskarpa miljögränser (in Swedish). Svensk Botanisk Tidsskrift. 104(1); 23‐31 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Paper I and appendix I and II are reproduced with permission from Wiley‐Blackwell (I), Nepper and 

Stagehøj (appendix I) and Svenska Botaniska Föreningen (appendix II)  

3

 

4

Content  

Content 

 

7      Abstract 

9      Resumé (Danish summary) 

11      Summary 

13      Introduction 

18      Aims 

19      Paper synopsis 

25      Conclusion and perspectives 

26      Acknowledgements 

27      References 

31      Paper I 

45      Paper II 

61      Paper III 

81      Paper IV 

103      Appendix I 

111      Appendix II 

122      Supplementary material 

 

 

6

Abstract 

Abstract 

A large variety of plant species of very different evolutionary origin are found within and along the 

margins of aquatic ecosystems. These species have very different adaptations depending on the 

particular environmental condition under which they grow. This thesis examines the role of these 

adaptations or functional traits for the distribution on large scales and along specific 

environmental gradients. Characean algae (charophytes) are an ancient group of aquatic plants 

found in most aquatic ecosystems. I confirmed that they have declined markedly during the 20th 

century, most likely as a consequence of widespread eutrophication, and that declining species 

were characterized by traits such as being obligate perennial and having preference for alkaline 

lakes. Partly as a consequence of this, there was an exceptionally high proportion of rare 

charophyte species in Denmark and Scandinavia. Theses rare species are specialists in particular 

environments, while the abundant species have traits such as broad salinity tolerance, tall shoots, 

vegetative reproduction and variable life form. Vascular plants, in contrast to charophytes, occupy 

the entire gradient from submerged to drained conditions. Along this gradient their functional 

traits vary markedly due to the environment conditions filtering away species lacking suitable 

adaptations. I found that the filtering of traits was highly dependent on specific traits and position 

along the gradient and the traits experiencing the strongest environmental filtering also proved to 

be the most important in predicting the species composition under local conditions. The results 

stress the importance of environmental filtering as mechanism determining community assembly 

in plant communities. The zone from submerged to drained conditions along the margins of 

softwater, oligotrophic lakes is often dominated by isoetids. They have the remarkable adaption of 

using sediment CO2 taken up through permeable roots. This trait makes isoetids vulnerable to 

sediment anoxia caused by degradation of organic sediments following eutrophication. I 

investigated if sediment desiccation leads to high mineralization without stressing the plants. The 

sediment mineralization increased following desiccation and plants continued to grow well under 

drained, fluctuating and submerged conditions stressing the potential use of desiccation as a 

sediment restoration tool in eutrophied systems. In conclusion, species functional traits have a 

large influence on distribution of species both on large scales and along specific environmental 

gradients. 

 

7

 

8

Resumé  

Resumé Naturlige ferskvandssystemer indeholder en lang række plantearter med meget forskellige udformning og evolutionær oprindelse. Disse planter tilhører en lang række familier af karplanter samt en enkelt familie af kransnålalger. Planterne spiller en afgørende rolle i akvatiske økosystemer, da de både har afgørende betydning for artsrigdommen af andre organismer, der benytter dem som føde eller skjul, og deres struktur og stofskifte påvirker det fysiske og kemiske miljø omkring dem. Diversiteten af planterne selv ændrer sig kraftigt langs gradienter i miljøforholdene på grund af de enkelte arters specialisering til mere eller mindre specifikke forhold. Arternes miljøkrav, deres niche, er et resultat af deres morfologiske og fysiologiske tilpasninger. Disse tilpasninger kaldes under et for funktionelle karaktertræk og har altså afgørende indflydelse på arternes tilstedeværelse og succes under forskellige miljøforhold. Denne Ph.D.‐afhandling omhandler planternes fordeling og karaktertrækkenes indflydelse på denne fordeling. 

De første to kapitler fokuserer på kransnålalger, der til trods for deres ofte betydelige effekt i de økosystemer, hvor de vokser, er dårligt undersøgte og ofte er oversete i forhold til de mere grundigt beskrevne karplanter. Kapitel I beskriver udviklingen af artsrigdom og hyppighed af kransnålalger i danske ferske vande siden 1940erne. Generelt har der være en nedgang i artsantallet i en række søer i løbet af 1900 tallet og tre arter er uddøde i Danmark, mens en enkelt ny er indvandret. Denne udvikling skyldes med stor sandsynlighed den kraftige forurening med næringsstoffer gennem perioden, som har medført en udbredt eutrofiering. For de enkelte art er mønsteret, at arter der før var almindelige, er blevet endnu mere almindelige, mens de sjældne arter er gået tilbage. De karaktertræk, der er typiske for de arter der er gået tilbage arter, er at de er flerårige og har forkærlighed for hårdvandede søer. Disse forhold er typisk for arter tilknyttet hårvandede klare og dybe søer, hvilket er en sjælden habitattype i nutidens Danmark. Et lyspunkt er dog, at artsrigdommen har en tendens til at stige i de seneste 20 år som følge af en reduceret næringsbelastning af søerne. 

I kapitel II beskrives udbredelsen af kransnålalger i de Nordiske lande og hvordan den enkelte arts nichespecialisering og karaktertræk har betydning for arternes hyppighed. Analysen viste at er der en meget stor andel af sjældne arter af kransnålalger (50‐87 %) også i forhold til karplanterne (30‐35 %). De sjældne og fåtallige arter er karakteriseret af specialiserede nicher og en lavere skudhøjde, mindre salttolerance, dårligere vegetativ formering samt en specialiseret livsform.  

Den enkelte arts karaktertræk kan have stor betydning for dens udbredelse under forskellige miljøforhold. Denne mekanisme er beskrevet som miljøfiltrering hvor miljøforholdene virker som en række filtre, der fjerner arter uden de rette træk. I kapitel III undersøgtes, hvilke træk, hos karplanter, der var mest filtreret af miljøforholdene langs en vand gradient. Analysen viste, at filtreringen af de enkelte karaktertræk generelt ændrede sig med miljøforholdene, mens nogle træk som rodporøsitet og det specifikke bladareal var udsat for filtrering langs hele gradienten. De mest filtrede træk var også afgørende for at kunne forudsige artssammensætningen under specifikke miljøforhold. Dette resultat viser empirisk at miljøfiltring er af afgørende betydning for artssammensætningen i plantesamfund. 

Vandgradienten fra permanent neddykkede til kun tidsvis vandækkede områder langs oligotrofe blødvandede søer, hvor vandstanden svinger, udgøres vegetation ofte af grundskudsplanter, især af planten strandbo Littorella uniflora. Grundskudplanterne er pga. deres gaspermeable rødder særligt sensitive over for dårlige iltforhold i sedimentet, hvilket ofte er 

9

Resumé  

forårsaget af øget organisk stoftilførsel som følge af eutrofiering. Da lufteksponerede sedimenter har en højere ilttilgængelighed og dermed højere omsætning af organisk stof undersøgte vi i kapitel IV om udtørring kan forbedre sedimentforholdene uden at stresse strandbo. Strandbo opnåede den højeste biomasse under drænede forhold og selv under vekslende vanddække, hvor planterne var tvunget til ar skifte mellem luft‐ og vandblade havde planterne en positiv vækst. Omsætningen målt som frigivelse af uorganisk kulstof var markant højere under drænede forhold. Søer bør derfor kunne fluktuere frit og øgede fluktuationer kan benyttes som restaureringsværktøj i situationer hvor søsedimenter er blevet uegnede for plantevækst – især af grundskudsplanter. 

I appendikset indgår et dansk og et svensk kapitel, begge udgivne i populærvidenskabelige tidskrifter. Det første kapitel omhandler alle danske arter af kransnålalger og deres udbredelse. Der gives en vurdering af arternes risiko for at uddø nationalt. Denne vurdering, en rødlistevurdering viser, at 85 % af de danske arter har en forøget risiko for at uddø, og at tre arter allerede med stor sandsynlighed er uddøde i Danmark. 

Det andet kapitel undersøger planternes fordeling, deres karaktertræk og artrigdom langs en gradient i jord og vanddybde. Langs gradienten var der en stor artsudskiftning, og arter der fandtes få meter fra hinanden havde markant forskellige karaktertræk som tilpasning til netop det mikromiljø, hvor de voksede. Som følge af et mere produktivt miljø og en større regional artspulje fandtes den største artrigdom ved stor jorddybde og god vandtilgængelighed. Arternes karaktertræk har altså meget stor betydning for arternes udbredelse og fordeling langs 

miljøgradienter. For vandplanterne og især kransnålalgerne kender vi dog fortsat kun et fåtal af 

trækkene og ved endnu mindre om deres rolle langs miljøets mange gradienter. 

10

Summary  

Summary Natural freshwater ecosystems contain a large variety of plants with very different structure and evolutionary origin. These species include representatives from a large number of vascular plant families and one family of characean algae, or charphytes. Plants play a pivotal role for ecosystems by supporting a large biodiversity of associated organisms and by their structure and metabolism affecting the chemical and physical environment. The diversity of the plants themselves changes drastically along environmental gradients as a consequence of species specialization to more or less specific habitat conditions. The particular requirements of a given species, the species niche, are the result of morphological and physiological adaptations. These adaptations are termed functional traits and have decisive influence on species success under particular environmental conditions. This Ph.D.‐thesis is about plant distribution and the influence of functional traits. 

The first two chapters focus on charophytes, which despite their extensive distribution and considerable ecosystem effects often have been overlooked in ecological studies. Chapter I analyses the changes in species richness and abundance of charophytes in Danish freshwaters during the 70‐years long period from about 1940 to 2010. Overall, the richness has declined in reinvestigated lakes and three species have gone nationally extinct during the period while one species has immigrated. This development is most likely due to the widespread eutrophication occurring throughout the 20th century as a result of greater anthropogenic input of nutrients to aquatic ecosystems. The pattern of single species is such that common species have become even more common and rare species have become even rarer. While species increasing in abundance tend to have wide tolerances to alkalinity and nitrogen content, the declining species has functional traits such as a perennial life form and a preference for alkaline lakes, typical of deep‐growing species in large alkaline lakes, a rare contemporary habitat. A positive trend in species richness of lakes in recent years, however, gives hope for the future. 

Chapter II analyses the distribution of charophyte species in Scandinavia and how niche specialization and functional traits influence species abundance. The analysis revealed that a high proportion of species are rare and Red‐listed (50 to 87 %) compared to aquatic vascular plants (30 to 35%). Rare charophyte species are characterized by a high degree of niche specialization and low shoot height, low salt tolerance, poor vegetative reproduction and a specialized life form.    Functional traits can have a profound impact on the distribution of particular species along environmental gradients. The mechanism has been described as environmental filtering where the environmental conditions act as a set of filters removing species without well suited traits. In chapter III, it was investigated to what extent functional traits of vascular plants experienced filtering along a hydrological gradient from below water in ponds to very dry conditions on shallow soils covering the limestone pavements on Øland’s alvar. The analysis showed that the filtering of single traits changed depending on the environmental conditions, while traits such as root porosity and specific leaf area were significantly filtered throughout the investigated gradient. The traits experiencing the strongest environmental filtering also proved to be the most important in predicting the species composition under particular conditions. The results stress the importance of environmental filtering as a mechanism driving community assembly in plant communities.    The hydrological gradient from permanently submerged to areas only occasionally flooded along the margins of oligotrophic, softwater lakes are often dominated by plants of the isoetid life form, particularly Littorella uniflora. Isoetids are particularly sensitive to organic sediment deposition 

11

Summary  

following eutrophication because of gas‐permeable roots and poor anchorage in organic unconsolidated sediments. As drained sediments have higher oxygen availability and, thus, a higher mineralization rate it was investigated in chapter IV if desiccation could improve sediment conditions without stressing L. uniflora. The species developed the highest biomass under drained conditions. The biomass also increased under fluctuating water levels, where the plants are inclined to change between aerial and submerged leaves, showing a high tolerance across the entire hydrological gradient. The sediment mineralization rate measured as efflux of inorganic carbon was markedly higher under drained conditions which also led to significantly more consolidated sediments both conditions potentially improving isoetid plant growth. Therefore, lakes should fluctuate freely and periods of sediment desiccation can be used as a restoration tool at locations where too high organic content has made lake sediments unsuitable for plant growth.  The appendix includes a Danish and a Swedish chapter dealing with popularization of science. The paper in Svensk Botanisk Tidskrift does, however, include original data. The Danish chapter considers all Danish species of charophytes and their distribution along with an assessment of the risk of species going nationally extinct, a Red List evaluation. The assessment showed that 85 % of the Danish charophyte species have an increased risk of going extinct nationally and three species have most likely already been lost from Denmark.   In the second appendix chapter in Swedish, the distribution of plants, their functional traits and the species richness were investigated along a gradient in water cover and soil depth on Öland’s alvar. Along the gradient, there was a large species turnover and species found within meters from each other had markedly different functional traits adapted to the particular micro‐environment where they were found. Species richness was highest in deep soils with good water supply as consequence of more a more productive environment and a larger regional species pool.    In conclusion, species functional traits have a large influence on the distribution of species both on large scales and along specific environmental gradients. The knowledge of charophyte distribution and traits, the prediction of species abundance through their traits and the effect of water levels on plant performance and sediment conditions all have potential consequences for ecosystem management. For aquatic vascular plants and particularly for the charophytes only a minority of traits have been identified and even less is known about their role along the multitude of environmental gradients. For vascular plants, improved understanding of how species traits are filtered by environmental conditions can potentially ensure more accurate future predictions of community assembly.  

12

Introduction  

 

Introduction 

Biodiversity of aquatic plants 

Natural  freshwater  systems  contain  a  large 

variety of plants of different growth form and 

phylogenetic  origin.  The  macroscopic  plant 

communities of  freshwaters  in  the Northern 

temperate  regions  include  representatives 

from  many  families  of  vascular  plants  and 

one  family of  characean algae  (Chambers et 

al.  2008).  The  presence  of  aquatic  plants  in 

lakes and ponds  support a  large biodiversity 

of  associated  organisms  and  plays  an 

important  role  for  the  ecological  state  of 

water bodies by stabilizing bottom sediments 

and  providing  shelter  for  zooplankton 

(Jeppesen  et  al.  1998).  The  diversity  of  the 

aquatic plants  themselves  responds  strongly 

to natural gradients in habitat conditions but 

also  to  anthropogenic  alterations  of 

environmental conditions (Sand‐Jensen et al. 

2000,  Vestergaard  and  Sand‐Jensen  2000). 

During  the  last  century,  aquatic  plant 

richness  has  declined  in many water  bodies 

of the  intensively used  landscapes of Europe 

because  of  habitat  deterioration  and 

destruction  (Lachavanne  et  al.  1992,  Sand‐

Jensen  et  al.  2000,  James  et  al.  2005).  The 

response  of  individual  species  to 

environmental conditions varies considerably 

resulting  in  distinct  patterns,  with  some 

species  being widespread  and  common  and 

others  rare.  These  patterns  of  commonness 

and rarity have been shown to be related to 

species  niche  specialization  with  generalist 

species  being  more  widespread  and 

abundant  than  specialists  (Thompson  et  al. 

1999, Gaston and Spicer 2001). 

The  mechanism  by  which 

environmental conditions affect the presence 

and  abundance  of  particular  species  can  be 

characterized as the environment acting as a 

set  of  filters,  only  letting  through  well 

adapted  species  with  suitable  functional 

traits (Keddy 1992, McGill et al. 2006, Shipley 

et  al.  2006).  A  functional  trait  is  a  well‐

defined  measurable  property  of  organisms, 

usually  measured  at  the  species  level  that 

strongly  influences  organismal  performance 

and,  thus, can be used comparatively across 

species  (McGill et al. 2006). Functional  traits 

are  thus  important  for  species  niche 

properties  such  as  specialization  and 

tolerance  to  various  environmental 

conditions (Hutchinson 1957, Violle and Jiang 

2009). 

In the following chapters I will give an 

introduction  to  plants  along  hydrological 

gradients  with  emphasis  on  the  highly 

understudied group of charophytes and also 

present  some  of  the  environmental  factors 

regulating presence and abundance of plant 

species. Firstly,  I will  introduce  the concepts 

of  the  species  niche  and  environmental 

filtering  of  species  traits  which  is  the 

conceptual  framework  ensuring  the 

understanding  of  species  response  to 

environmental gradients. 

 

Species niche 

A species’ niche was historically described as 

an  n‐dimensional  space  where  each 

dimension  represents  one  range  in  abiotic 

conditions under which a species can survive 

and  reproduce  (Hutchinson  1957).  The 

conditions under which a species can survive 

without  the  presence  of  other  species  is 

termed  the  fundamental  niche  while  the 

conditions under which the species is actually 

found  in nature  is referred to as the realized 

niche  (Hutchinson 1957). Often of  interest  is 

13

Introduction  

 

the breadth (or width) of a species niche and 

its  location  compared  to  niches  of  other 

species  (Thompson  et  al.  1999,  Gaston  and 

Spicer 2001). The location of niche compared 

to  the  average  environmental  conditions  in 

the area in focus can be viewed as a measure 

of niche  specialization where  species having 

niches  far  from  the  mean  conditions  are 

more  specialized  than  species  with  niches 

close  to  the  average  conditions  (Doledec  et 

al. 2000). Species with specialized niches are 

often  found  to  be  rarer  than  species  with 

more  general  niches;  a  pattern  existing  if 

species that have general niches and are able 

to  use  a  wide  variety  of  resources  will 

achieve  high  local  densities  and  be  able  to 

survive  in  more  places  over  larger  areas 

(Gaston  and  Spicer  2001).  Specialist  species 

have  been  found  to  decrease  in  abundance 

compared  to  generalists  as  a  result  of 

historically  increasing  anthropogenic  impact 

in  both  terrestrial  and  aquatic  ecosystems 

(McKinney and  Lockwood 1999, Clavel et al. 

2011).  

 

Environmental filtering through species traits 

Because  of  niche  differences  between  plant 

species, the environmental conditions can be 

viewed  as  a  set  of  filters  sorting  which 

species  can  occur  in  a  given  community 

(Keddy 1992). The niche is an expression of a 

species  tolerance  along  different 

environmental axes and these tolerances are 

regulated  by  the  morphological  and 

physiological  characteristics,  or  functional 

traits,  of  the  species.  The  environmental 

filtering can  thus be viewed as operating on 

species  functional  traits  rather  than  species 

themselves  (McGill et al. 2006, Shipley et al. 

2006).  Along  environmental  gradients,  the 

average and variability of the traits expressed 

in  communities  change  and  these  changes 

are  associated  with  shifting  strength  of 

environmental filtering (Cornwell and Ackerly 

2009, Bernard‐Verdier et al. 2012). A smaller 

range  in  trait  values  than  expected  from by 

random –  trait clustering – may be  taken as 

an  indication  of  trait  filtering,  because 

species  having  trait  values  outside  a  given 

allowed  trait  range  succumb  (Keddy  1992, 

Cornwell  and  Ackerly  2009).  The  great 

advantage of using  functional  traits  and not 

species entities is that traits respond similarly 

to  environmental  gradients  across  large 

geographic  regions  thus  increasing  our 

knowledge on more general patterns (Diaz et 

al. 2004, Wright et al. 2004). 

The  recent  theoretical  developments 

regarding  traits  and  environmental  filtering 

have not  yet been  fully  implemented  in our 

understanding  of  vascular  plant  distribution 

along  hydrological  gradients.  For  aquatic 

vascular plants in general and charophytes in 

particular, the knowledge of the  influence of 

traits on species tolerance and distribution is 

very  limited  (Blindow  1992,  Sand‐Jensen  et 

al. 2000). 

 

Charophytes and aquatic vascular plants 

The vascular plants and charophytes we meet 

today originate from single celled green algae 

living  700‐800 million  years  ago  (Butterfield 

et  al. 1988).  These unicellular  algae evolved 

into  multicellular  organisms  eventually 

yielding  the  group  of  charophytes  already 

before  the  Silurian  period  more  than  450 

million  years  ago  (Martin‐Closas  2003).  The 

charophytes are the closest living relatives to 

the  vascular  plants  dominating  especially  in 

terrestrial  ecosystems  (Karol  et  al.  2001). 

14

Introduction  

 

Vascular plants are, however, also  important 

representatives of most aquatic  systems but 

this  group  is  thought  to  have  evolved  from 

land  plants  re‐entering  the  aquatic 

environment  the  last  120  million  years 

(Sculthorpe  1967).  It  is  noteworthy,  that 

charophytes  dominated  shallow  freshwater 

and  coastal marine waters during  the entire 

preceding period  from 450‐120 million years 

ago. 

The  charophytes  of  today  are  5‐200 

cm  tall  macro‐algae  living  attached  to  soft 

sediments by rhizoids in both freshwater and 

marine  environments  (Krause  1997). 

Although  their  appearance  is  similar  to 

several  aquatic  vascular  plants  their 

morphology  is  very  different.  Charophytes 

consist  of  single  extraordinarily  long 

internodal  cells  connected  by  shorter  nodal 

cells  in an axial section from which branches 

radiate  above  the  sediment  and  rhizoids 

below  the  sediment  surface.  Contrary  to 

charophytes  the  vascular  plants with which 

they  share  habitat  have  highly  specialized 

multicellular  tissues  enabling  fast  internal 

transport  of  nutrients,  photosynthates  and 

gases  (Pedersen  and  Sand‐Jensen  1993, 

Raven et al. 2003). 

 

Functional traits of plants along gradients 

Despite  the  morphological  differences, 

charophytes  and  aquatic  vascular plants  are 

found  in  the  same habitats  and  share many 

functional  adaptations  or  traits  (Fig.  1). 

Among the most severe restrictions to plants 

living  submerged  in  water  is  the  104‐fold 

lower gas diffusion coefficient  in water  than 

in air potentially limiting the supply of CO2 for 

photosynthesis. In water it is particularly  

advantageous  to  have  a  large  surface  to 

volume ratio  to  increase uptake of dissolved 

ions  and  gases  and  aquatic  vascular  plants 

often have thinner and more serrated  leaves 

than  their  terrestrial  counterparts  while 

charophytes  have  notoriously  thin 

photosynthetic tissues because of the single‐

celled basic structure. Several species of both 

charophytes  and  aquatic  vascular  plants 

compensate  for  the  low  CO2‐concentration 

by  actively  using  dissolved  HCO3‐  found  in 

much  higher  concentrations  in  alkaline 

waters  than  dissolved  CO2  (Madsen  and 

Sand‐Jensen  1991,  Maberly  and  Madsen 

2002).  

 

 Fig. 1. Charophytes growing together with the vascular 

plant Littorella uniflora in a shallow lake on the island 

of Öland, Sweden. Photo: Ole Pedersen. 

15

Introduction  

 

In  water,  light  attenuation  is  often 

high  and  it  can  be  greatly  increased  by 

particles  or  dissolved  colored  compounds 

(Kirk  1994).  A  trait  that  can  partly 

compensate  for poor  light  condition  is a  tall 

upright  structure  allowing  leaves  to  grow 

close to the surface and tall species are thus 

favored under turbid conditions (Sand‐Jensen 

et al. 2008). 

While  charophyte  species  are  only 

capable  of  growing  under  water,  vascular 

plants are  found all along  the gradient  from 

submerged  to  dry  conditions.  The  traits  of 

vascular plant species distributed at different 

parts  of  the  gradient  tend  to  differ 

predictably  because  of  the  strong  abiotic 

selection  (Violle et al. 2011, Bernard‐Verdier 

et al. 2012). Traits  such as  specific  leaf area 

(SLA) change along hydrological gradients as 

plant  species  living  on  moist  soils  tend  to 

have  thin  leaves  with  a  large  surface  area 

while  species  living  under  drier  conditions 

have small and thick leaves to limit water loss 

(Bernard‐Verdier et al. 2012). 

 

Environmental conditions  

One  of  the  main  environmental  drivers  in 

aquatic  ecosystems  is  eutrophication 

triggered  by  high  nutrient  concentrations 

from  large  anthropogenic  inputs  of  waste 

water  and  farmland  fertilizers  (Moss  1998, 

Wetzel 2001). In eutrophied waters, the high 

nutrient  concentration  increases  growth  of 

both  microscopic  and  filamentous  algae 

impoverishing  the  light  conditions  and 

leading  to  increased  organic  matter  in 

sediments  as  the  algae  die  and  sink  to  the 

bottom  (Kalff  2002).  Decreased  light 

availability  reduces  the  bottom  area  with 

sufficient  light  available  for  plant  growth 

while  organically  enriched  sediments 

negatively affects root development and root 

anchorage  in  the  sediment  (Schutten  et  al. 

2005,  Raun  et  al.  2010).  Eutrophication  has 

thus  caused  decreased  species  richness  of 

both charophytes and aquatic vascular plants 

and  in  the  most  severe  instances  led  to 

complete  loss  of  submerged  vegetation  (ie. 

Lachavanne  et  al.  1992,  Sand‐Jensen  et  al. 

2000, Sand‐Jensen et al. 2008). As a result, a 

large  proportion  of  aquatic  vascular  plants 

(30‐35%) and particularly charophyte species 

(50‐87%) have entered the national Red Lists 

in  Scandinavia  due  to  their  high  risk  of 

extinction  (Gärdenfors  2010,  Kålås  et  al. 

2010,  Koistinen  2010, Wind  and  Pihl  2010). 

The  decline  of  abundance  due  to 

eutrophication is not the same for all species 

because  some  are  more  tolerant  to  the 

poorer  light  and  changed  sediment 

conditions  than  others  (Sand‐Jensen  et  al. 

2008). An anthropogenic disturbance such as 

eutrophication  can  lead  to  homogenization 

of  the  plant  communities  with  most 

communities  dominated  by  a  few  tolerant 

species  (McKinney  and  Lockwood  1999). 

While  the eutrophication effects on vascular 

aquatic plants are well documented, the long 

term consequences of eutrophication for the 

highly threatened charophytes have not been 

studied  in  detail.  It  is  therefore  of  great 

importance  to  describe  the  effects  on  this 

group  of  organisms  and  investigate  which 

species  traits are  involved  in regulating both 

the  overall  species  abundance  and  the 

response  to  specific  stressors.  These 

questions are  the  focus of papers  I and  II  in 

my thesis. 

Another  important  environmental 

factor  along  the  margins  of  freshwater 

16

Introduction  

 

systems is the location and fluctuation of the 

water  table  driven  by  seasonal  climatic 

variations  in  natural  systems  or  by 

impoundment in impacted water bodies (van 

der  Valk  2005,  Mjelde  et  al.  2013).  The 

change  of  hydrological  conditions  from 

permanently  submerged  to  dry  conditions 

plays  a  major  role  in  individual  species 

distribution  and  the  species  richness  along 

hydrological  gradients  (Silvertown  et  al. 

1999).  Changes  are  caused  by  each  species 

being, to some degree, specialized to certain 

hydrological conditions,  i.e. having a  specific 

hydrological  niche.  This  specialization  is  a 

function  of  the  traits  possessed  by  a  given 

species.  A  better  understanding  of  the 

species  distribution  patterns  along 

hydrological gradients can thus be gained by 

investigating  which  functional  traits  are 

important  in  determining  the  community 

assembly  at  particular  hydrological 

conditions. This  is  the aspect  is addressed  in 

paper III. 

In  the  zone  of  the  water  level 

fluctuation aquatic plants and sediments are 

exposed  to  both  submerged  and  drained 

conditions.  These  alterations  require  plant 

species with  specific  traits  enabling  survival 

under  both  conditions  such  as  leaf 

dimorphism  and  drought  resistant  seeds 

(Robe and Griffiths 1998, Brock et al. 2003). 

Fluctuating water  levels  also  have  profound 

implications  for  the  sediment  chemistry and 

mineralization  (Baldwin  and Mitchell  2000). 

Plants  with  an  amphibious  life  form  and 

being able to change their leaf morphology as 

a  consequence  of  the  changing water  cover 

are often benefitted by fluctuating conditions 

(Farmer  and  Spence  1986,  Casanova  and 

Brock  2000).  In  sediments,  lowering  of  the 

water  level  greatly  increases  the  oxygen 

availability  potentially  increasing  the 

mineralization  rate  of  sediment  organic 

matter  (Degroot and  van Wijck 1993,  James 

et  al.  2001).  Desiccation  can  also  increase 

sediment  consolidation  (James  et  al.  2001). 

Fluctuating  water  levels  can  thus  to  some 

degree counteract the chemical and physical 

stresses  associated  with  increased  organic 

sedimentation  following  eutrophication. 

Desiccation and subsequent  re‐submergence 

have however, also been  shown  to  result  in 

large  nutrient  fluxes  from  sediments  and 

substantial  acidification  due  to  oxygenation 

of  nutrient‐rich  organic  substances  and 

metal‐sulphides  (James  et  al.  2001,  Van 

Wichelen  et  al.  2007). Despite  the potential 

downsides, water  level  fluctuation has been 

suggested  as  a  tool  to  restore  sediment 

conditions and  improve plant growth  (Coops 

and  Hosper  2002).  In  paper  IV  we 

investigated  how  water  level  fluctuations 

affected  plant  growth  and  sediment 

conditions  in  isoetid populations particularly 

vulnerable  to  organic  enrichment  (Sand‐

Jensen  et  al.  2005, Møller  and  Sand‐Jensen 

2011). 

 

 

 

 

 

 

 

 

 

 

 

17

Aims  

  

Aims 

The overall aim of this thesis was to examine 

the  species  distribution  and  the  importance 

of  species  functional  traits  for  charophytes 

and  vascular  plants  in  water  and  along 

hydrological  gradients.  Particular  focus  was 

on  analysing  the  role  of  traits  and  niche 

characteristics  in  shaping  the  present 

distribution  of  charophytes  in  Scandinavia 

and  how  the  abundance  has  changed  in 

Denmark  during  the  last  century.  Emphasis 

was  also  given  to  determining  the 

importance  of  in‐lake  ecological  conditions 

and  landscape  variables  for  the  changes  in 

species  richness  following  a  century  long, 

anthropogenically  imposed,  large  scale 

alteration of the Danish landscape. 

Along  hydrological  gradients,  the 

objective was  to  investigate  to what  extent 

specific traits are subjected to environmental 

filtering  and  if  traits  experiencing  filtering 

were  important  in  predicting  community 

assembly.  The  aim was  also  to  describe,  in 

detail, the response of a single plant species 

(e.g.  Littorella uniflora) and  the  sediment  to 

differing  hydrological  conditions.  As 

desiccation  can  lead  to  improved  sediments 

conditions  for  plants,  focus  was  on  the 

potential role of water level fluctuations as a 

restoration tool. 

   

 

 

The data background  for  the analyses was a 

combination of data available in databases or 

in  published  literature  and  field 

investigations  combined  with  experiments 

performed in the laboratory.  

  By  consolidating  the  use  of  species 

functional  traits  into  aquatic plant  ecology  I 

anticipate  to  gain  a better understanding of 

the processes  regulating  species distribution 

and abundance. I also hope that the obtained 

knowledge  can  be  of  use  to  authorities  by 

improving  management  of  aquatic 

ecosystems.  

 

 Fig. 2. Analysing the vegetation and collecting samples 

for  trait  analysis  near  a  desiccatin  pond  on  Öland, 

Sweden. Photo: Hans Henrik Bruun. 

 

 

 

18

Paper synopsis  

  

Paper synopsis Seventy years of changes in the abundance of Danish charophytes (paper I) Charophytes  grow  attached  to  soft  bottoms in  ponds,  streams,  lakes  and  estuaries.  Like the aquatic vascular plants, charophytes have been  reported  to  decline  during  the  20th century, but  the  causes  and  species  specific changes  are  poorly  understood  (Blindow 1992, Simons and Nat 1996, Auderset Joye et al.  2002).  We  used  Danish  studies  on freshwater  charophyte  distribution conducted  around  1940  (Olsen  1944)  and repeated  measurements  performed  by  the national  monitoring  programme  and ourselves during recent years to evaluate the historical  development  of  species  richness and  dominance  patterns. We  also  tested  to what  extent  historical  changes  of  species abundance  in  29 water  bodies were  related to  landscape  features,  water  quality  and species traits. 

We  found  that  three  species  of freshwater  charophytes  (Chara  filiformis, Tolypella  intricata  and  Nitella  gracilis)  have apparently disappeared  from Denmark while one  species  (Chara  connivens)  has immigrated. The national species  richness of freshwater  charophytes  has  thus  declined from 21 to 19 species. 

Species  abundance  based  on occurrence  in many water bodies  followed a linear rank – log abundance relationship both in  the historical  and  the  recent  studies  (Fig. 3).  The  dominance  structure  was  stronger today  than  historically  as  common  species have  become  relatively more  abundant  and uncommon species relatively rarer (Fig. 2).  

Among  species  traits,  perennial  life form  and  preference  for  alkaline  waters typical  of  deep‐growing  species  in  large alkaline  lakes,  a  rare  contemporary  habitat, were  significantly  related  to  the  historic species  decline.  Species  increasing  in 

abundance had wide  tolerances  to alkalinity and water nitrogen content. Twenty‐nine  lakes  and  ponds  studied 

repeatedly  showed  a  significant  decline  of 

mean species richness from 3.4 to 2.4 during 

the  70  years.  A  small  increase  of  species 

richness  has  taken  place  during  the  recent 

15‐20  years  in  several  lakes  experiencing 

reduced  nutrient  loading.  However,  many 

species  survive  today  in  relict  populations 

and may find it difficult to recolonize lakes in 

which  water  quality  has  improved.  The 

historical  decline  of  species  richness  was 

significantly  related  to  higher  nutrient 

concentrations,  higher  phytoplankton 

biomass  and  lower  transparency  of 

eutrophied  water  bodies.  In  contrast,  the 

amount  of  wetlands  and  openness  of  the 

landscape  close  to  a  water  body  did  not 

predict  the  historical  development  perhaps 

because  local  processes  or  long  distance 

dispersal  determine  species  richness. 

0 5 10 15 20-0.5

0.0

0.5

1.0

1.5

2.0

Rank

log

(abu

ndan

ce)

 Fig.  3.  Abundance  vs.  species  rank  (x‐axis)  of  the 

fifteen most abundant Danish freshwater charophytes 

in  1940  (open  circles)  and  in  2010  (closed  boxes). 

Abundance was calculated as log (x + 1), where x is the 

%  species  frequency  of  occurrence.  The  linear 

regression  line  for  the  fifteen most abundant  species 

was significantly steeper in 2010 than 1940 (p < 0.01). 

 

19

Paper synopsis  

  

Considering  together  the  loss  of  former freshwater habitats and  the deterioration of surviving  habitats,  we  conclude  that charophyte occurrence has declined by about 56%  in  Denmark  during  the  last  70  years. Species  reductions  determined  from reinvestigated  water  bodies  (29%)  and  the national  species  list  (10%)  are  much  lower and  are  less  suitable  measures  of  the developmental status of charophytes. 

 Niche  specialization  and  functional  traits regulate  abundance  of  Scandinavian charophytes (paper II) Given  the  historic  loss  of  charophyte abundance,  eutrophication  of  their  habitats and their competitive inferiority to tall rooted plants  under  nutrient  rich  conditions  (Van den Berg et al. 1998), it should be anticipated that  charophytes  are  among  the  most threatened  phototrophs.  Because  species with  specialized  niches  are  by  definition restricted  to a  few  special, but very  suitable habitats  they  are  suggested  to  remain  rare overall  because  the  majority  of  habitats would  be  unsuited  for  them  (Brown  1984, Clavel et al. 2011). We thus hypothesize that within  the  charophyte  group,  species  with specialized  niches  and  functional  traits  such as low salinity tolerance and low shoot height should  be  restricted  because  of  few  growth habitats and low competitive capability. 

This  prediction  was  tested  by examining  the distribution of charophytes  in Scandinavia  and  the  relationship  between species  abundance  and  species  traits. Abundance  values  were  assigned  to  each species based on their status on the national Red Lists. The specialization of species niches were evaluated by  the Outlying Mean  Index (OMI,  Doledec  et  al.  2000)  using  a  large Danish  dataset  on  charophyte  distribution and  environmental  variables  while  species functional  traits  were  derived  from  the literature.  

The  results  supported  our  predictions showing that 50‐87% of the species were on the  national  Red  Lists  in  the  four Scandinavian  countries.  These  proportions are much  greater  than  of  aquatic  (30‐35%) and  terrestrial  plants  (18‐28%).  Species abundance  of  charophytes  in  Scandinavia decreased  significantly  with  niche specialization  in separate analyses of marine and  freshwater sites  (Fig. 4). Four  functional traits:  shoot height,  salinity  tolerance, bulbil production  and  ability  to  express  both perennial  and  annual  life  form  were significantly  positively  related  to Scandinavian abundance.   

0

20

40

60

80

100

Fre

shw

ater

OM

I (%

)

0 10 20 300

20

40

60

80

100

Scandinavian abundance

Mar

ine

OM

I (%

)

Fig.  4.  The  relationship  between  niche  specialization and Scandinavian abundance. The niche specialization was calculated as the Outlying Mean Index (OMI) using specific  environmental  conditions  for  Danish  growth sites (a large OMI indicate a high speciealization). The niche  specialization  was  calculated  based  on  two datasets,  using marine  (>0.5  g NaCl  L‐1)  growth  sites and  freshwater  sites.  The  lines  shown  are  model  II regressions.  The  relationships  between  abundance and OMI are significant (Spearman, p< 0.05). 

20

Paper synopsis  

  

We  conclude  that  charophytes are generally rare  species  restricted  by  habitat  loss  and competition  with  fast‐growing phytoplankton, macroalgae  and  tall  vascular plants at eutrophic sites. The few charophyte species  maintaining  a  high  abundance  in Scandinavia are generalists tolerant of a wide range  of  conditions,  while  specialists  are often  short,  have  a  restricted  life  form  and are  rare  in  the  disturbed  contemporary landscape.  

From pond  to alvar: predicting plant  community composition  along  a  steep  hydrological  gradient (paper III) 

Conceptually,  community  assembly  may  be viewed  as  a  sequence  of  filters,  sieving  off species  according  to  their  functional  traits and  thus  restricting  the  range of  trait values within  a  community  (Keddy  1992,  Cornwell and  Ackerly  2009).  This  conceptual framework may  aid  in  achieving  one  of  the core  aims  to  ecology:  the  prediction  of species’  occurrence  and  abundance  from knowledge  of  the  environmental  conditions (McGill et al. 2006, Shipley et al. 2006). One suggested  path  towards  this  goal  is  the maximum  entropy  (maxent)  model developed  to predict  species’ abundances  in local  communities  (Shipley  et  al.  2006, Shipley  2010).  The  strength  of  filtering  of particular  traits  along  environmental gradients is, however, still not clarified and it is also not known  if  filtered  traits are better predictors of community assembly. 

We  therefore  investigated  plant community  composition  and  six  functional traits  along  a  pronounced  hydrological gradient  and  studied  the  strength  of environmental  filtering  on  traits  and  the importance  of  the  different  traits  to community assembly. 

The  results  showed  that  the functional  traits  expressed  by  plant 

communities  change  along  the  hydrological gradient  and  that  the  range  of  trait  values observed was restricted at particular parts of the  gradient  indicating  environmental filtering (Fig. 5). Further analysis showed that root  porosity  is  significantly  filtered  by  the environment along the gradient, as is specific leaf  area  and  resistance  to  water  loss  on drying.  For  individual  traits,  the  strength  of filtering  wax  and  wane  along  the  gradient, strongly  suggesting  that  the  mechanisms through  which  species  are  filtered  into communities  act  through  different  traits  as environmental conditions change.   The maxent model predicted 66% of the variation  in  species’  relative abundances using  six  traits. The  traits  subject  to  filtering also  were  most  important  in  predicting species abundance. 

In  summary  we  showed  that  few plant  traits  are  subject  to  environmental filtering  across  the  entire  gradient  in water coverage and soil depth and that most traits experience strong filtering in only part of the gradient.  Evidence  for  congruence  between trait  dispersion  indices  and  the  maxent model  was  established,  underpinning  the importance  to plant community assembly of environmental  filtering  of  species  through their traits.  

Water  level  fluctuations  affect  plant  growth and sediment dynamics of isoetid populations in oligotrophic lakes (paper IV) Amphibious species growing in the zone of 

water level fluctuations often have traits 

allowing them to cope with strong abiotic 

changes accompanying alterations from 

drained to submerged conditions. Along the 

margins of oligotrophic softwater lakes the 

isoetid species, Littorella uniflora, survives 

this stress by changing leaf morphology upon 

emergence (Hostrup and Wiegleb 1991, Robe 

21

Paper synopsis  

  

 Fig. 5. Community‐weighted mean trait values (CWM, black dots) and ranges of trait values (bars) in local communities along a hydrological gradient  (expressed as PC1) of  increasingly drier conditions. The curves  represent  least square regression models and r2‐values indicate strength of model fit. Linear regression were used for leaf dry matter content (LDMC) while second order polynomial  regression  lines were  fitted  to  specific  leaf area  (SLA), water  loss on drying (WLOD), plant height, leaf area (log), and root porosity.   and Griffiths 1998) and it may be particularly successful  because  of  its  ability  to  take  up CO2  across  root  surfaces  under  both submerged  and  drained  conditions  (Fig.  6, Nielsen et al. 1991).   Accumulation  of  labile  organic matter  in the sediments from phytoplankton blooms  accompanying  eutrophication represents  a  threat  to  isoetids  because  of insufficient  root  oxygenation,  reduced anchorage  and  enhanced  growth  of  taller competitors. We hypothesized that L.  

 uniflora  tolerates  different water  levels  but grows better and maintains dominance when sediments  are  periodically  drained  because air  exposure  assists  oxygenation, mineralization  and  consolidation  of sediments.  If  so,  drainage  offers  a  possible restoration tool. We tested this hypothesis by measuring  plant  performance  and  sediment metabolism  under  drained,  fluctuating  and submerged  conditions  in  the  laboratory  and along elevation gradients in the field in  

0

200

400

600r2= 0.44

SLA

(cm

2 g

-1)

-6 -4 -2 0 2 40

10

20

30

40r2= 0.59

PC1

Roo

t por

osity

(%

)

0

100

200

300

400 r2= 0.31

LDM

C (

mg

dw/g

fw

)

0

20

40

60

80

100r2= 0.41

WLO

D (

%)

-6 -4 -2 0 2 4-2

-1

0

1

2

3r2= 0.65

PC1

log

(leaf

are

a, c

m2 )

0.0

0.5

1.0

1.5

2.0r2= 0.55

Hei

ght (

m)

22

Paper synopsis  

  

 organically  enriched  sandy  sediments  from an oligotrophic Danish lake.   Laboratory  and  field  experiments confirmed  that  L.  uniflora  maintained  high leaf chlorophyll and photosynthesis and grew in  biomass  under  all  conditions.  Increase  of biomass and shoot weight was highest under drained conditions and most so for roots as a suitable  response  to  higher  sediment oxygenation  and  leaf  transpiration.  The increase  of  biomass  was  lowest  under fluctuating  conditions  as  a  logical consequence  of  higher  costs  of  producing aerial  leaves  to  replace  unsuitable  aquatic leaves  upon  changes  from  submerged  to drained conditions. Efflux  of  inorganic  carbon  from mineralization  in  un‐vegetated  standard sediments  increased  5.5‐fold  from submerged,  over  fluctuating  to  drained treatments.    

 Fig.  6.  Emerged  Littorella  uniflora  forming  dense vegetation  along  the  margin  of  an  oligotrophic softwater lake. Photo: Ole Pedersen.  

 In  submerged  vegetated  sediments,  where part  of  sediment  CO2  is  incorporated  into  a growing plant biomass,  the sum of  inorganic carbon  efflux  and  plant  growth  had  more than  doubled  compared  to  bare  sediments probably  reflecting  stimulated  sediment oxygenation  and  mineralization  because  of root  oxygen  release.  Under  drained,  well‐oxygenated conditions the increase of carbon in  the  biomass  was  accompanied  by  an equivalent  decline  of  the  inorganic  carbon efflux  relative  to  bare  sediments.  Still,  the sum of inorganic carbon efflux and additional carbon  trapped  in  the  biomass  was  2‐fold higher  under  drained  than  submerged conditions.   The  proportion  of  reduced  Fe2+  to total Fe as a measure of the oxygen status of sediments  declined  from  submerged,  over fluctuating to permanently drained sediment. Also  the  proportion  of  reduced  Fe2+  was lower  in plant  covered  than bare  sediments reflecting root release of O2. Drainage and air contact also  increased  sediment density and consolidation  and,  thereby,  improved  root anchorage  compared  to  submerged conditions.    Periodic  drainage  and  air  exposure of  isoetid  covered  sediments  is  a  natural phenomenon  in  oligotrophic  seepage  lakes that  should  be  reinstalled  as  a  possible restoration  tool  if  the  water  has  been regulated  to  a  constant  level.  Water  level fluctuations  of  a  moderate  frequency  can benefit  isoetid  populations  because  they grow  and  survive  better  than  competitors and potential harmful organic matter can be mineralized without sediment anoxia.  Appendices In  the appendix are a Danish and a Swedish article.  The  first  addresses  all  the  Danish species of charophytes and their distribution along  with  an  assessment  of  the  risk  of 

23

Paper synopsis  

  

species  going  nationally  extinct,  a  Red  List evaluation. Danish  charophytes are  found  in most types of freshwater through brackish to fully marine environments. The knowledge of the  distribution  of  species  found  in marine and  larger  freshwater  habitats  are  fairly good,  but  the  distribution  of  species with  a preference  for  habitats  such  as  ponds  and fens  is  not  well  known.  The  assessment  of species  extinction  risk  showed  that  85 %  of the  Danish  charophyte  species  has  an increased risk of going extinct nationally and three  species have most  likely  already been lost from Denmark. 

In  the  second  appendix  chapter  the distribution  of  plants,  their  functional  traits and  the  species  richness  were  investigated along  a  gradient  in  water  cover  and  soil depth  in an abandoned  lime stone quarry on Öland. Along the gradient, there was a  large species  turnover  and  species  found  within meters  from  each  other  had  markedly different  functional  traits  adapted  to  the local  micro‐environment  where  they  were found. Some of the most extraordinary plants found was the colonial cyanobacteria Nostoc commune surviving on bare limestone   

pavements as dry crusts only to be revitalized by  rain  events  and  winter  flooding  (Fig.  7) and  several  charophyte  species  totally dominating  even  desiccating  pond  and  re‐sprouting from spores upon re‐submergence. The species richness was highest at high soil depth  and  good  water  supply  as  a consequence  of  more  a  more  productive environment  and  a  richer  regional  species pool.  

 Fig. 7. Typical vegetation on the limestone pavements in  the  investigated  quarry  on  Öland.  The  colonial cyanobacteria Nostoc commune and the vascular plant Sedum album surviving the tough conditions between submergence  in  winter  and  complete  dessication  in summer. Photo: Ole Pedersen. 

 

24

Conclusions and perspectives  

 

Conclusions and perspectives 

Several  concrete  results, new questions  and 

possible applications have emerged from this 

Ph.D.‐thesis. Concerning charophytes, I found 

that they were indeed rare and many species 

are  declining  in  abundance  in  the 

contemporary  Scandinavian  landscape.  The 

results  of  the  influence  of  traits  and  niche 

specialization on abundance and distribution 

could hopefully inspire further research to fill 

the  gap  in  our  knowledge  of  charophyte 

ecology. For most species, data on their basic 

biology stem from anecdotal studies most of 

them  performed  more  than  half  a  century 

ago  and  for  quite  a  few  species  new 

molecular  methods  have  questioned 

previously accepted taxonomical entities. It is 

thus urgent to gain a better understanding of 

the basic ecology, taxonomy and distribution 

of  species  in  small  water  bodies  such  as 

ponds, fens and bogs from which there is less 

data  available.  This  knowledge  is  of  pivotal 

importance  to  ensure  that  the  proper 

management can be implemented to prevent 

that  rare  and  declining  species  are  not  lost 

entirely. 

 

Because  knowledge  of  vascular  plants  is 

much more  advanced,  it  is  now  possible  to 

make  predictions  about  the  abundance  of 

these species along environmental gradients. 

I  have  provided  a  better  understanding  of 

how  species  traits  are  filtered  by 

environmental conditions and provided a link 

potentially  ensuring  more  accurate  future 

predictions  of  community  assembly.  Fully 

implemented,  this  knowledge  can  lead  to 

well‐considered  decisions  in  ecological 

restorations,  because  effects  of  a  given 

intervention  can  be  accurately  predicted 

beforehand. 

  One  such  restoration  could  be  the 

use  of  sediment  desiccation  of  lake 

sediments  that  proved  to  have  the 

prospective  to  improve  conditions  for  plant 

growth where  these have been deteriorated 

by  organic  enrichment.  Eutrophication 

leading  to  organic  enrichment  has  been 

widespread along with  stabilization of water 

levels  to  ensure  human  control  over  the 

environment,  however  letting  go  of  control 

would  lead  to  a  richer  and  more  diverse 

nature. 

 

25

Acknowledgements  

Acknowledgments 

I am indebted to numerous people supporting me through the last three and a half years. Firstly I 

want to thank Tone for bearing with me all these years and for helping me through the periods 

where everything looked impossible with strong help from Alfred’s smiles. To my supervisor Kaj, I 

am deeply grateful for your invaluable help, guidance and prompt feedback. You always take the 

time needed to get your students in the right direction and that is a great personal trait. I have 

really enjoyed our numerous fruitful discussions on plants, birds and nature in general, it is a 

pleasure to know a man that knows (almost) everything. To Claus, with whom I have shared path 

with since our first days of introductory chemistry even before the first semester in biology, thanks 

for the many years. I know we will keep hanging out and go fishing (we still lack the big one), but 

hopefully we can keep on collaborating scientifically as well. Big thanks to Hans Henrik, for 

introducing me to the fascinating world of terrestrial botany and for always having an open door 

and mind for a young aquatic ecologist. Thanks to Jens, for your always open ears and for calling 

me a bastard (skrummel) in the morning, to Ole P, for help, in particular with the pictures of this 

thesis, to Lars Iversen, for valuable statistical assistance, to Birgit, Ayoe and Anne, for skilful 

laboratory assistance, and to Mikkel, for help in the field even though the data did not get 

published (yet).  

Last but not least to my fellow young scientists at FBL: Anders, Ane, Jesper, Matteo, Theis, Laci, 

Nina, Frandsen and all the students thanks for making life as a PhD‐student a joy. 

 

26

References 

  

References Auderset Joye, D., E. Castella, and J. B. Lachavanne. 

2002. Occurrence of Characeae in Switzerland over the last two centuries (1800‐2000). Aquatic Botany 72:369‐385. 

Baldwin, D. S. and A. M. Mitchell. 2000. The effects of drying and re‐flooding on the sediment and soil nutrient dynamics of lowland river‐floodplain systems: A synthesis. Regulated Rivers‐Research & Management 16:457‐467. 

Bernard‐Verdier, M., M.‐L. Navas, M. Vellend, C. Violle, A. Fayolle, and E. Garnier. 2012. Community assembly along a soil depth gradient: contrasting patterns of plant trait convergence and divergence in a Mediterranean rangeland. Journal of Ecology 100:1422‐1433. 

Blindow, I. 1992. Decline of charophytes during eutrophication ‐ comparison with angiosperms. Freshwater Biology 28:9‐14. 

Brock, M. A., D. L. Nielsen, R. J. Shiel, J. D. Green, and J. D. Langley. 2003. Drought and aquatic community resilience: the role of eggs and seeds in sediments of temporary wetlands. Freshwater Biology 48:1207‐1218. 

Brown, J. H. 1984. On the relationship between abundance and distribution of species. American Naturalist 124:255‐279. 

Butterfield, N. J., A. H. Knoll, and K. Swett. 1988. Exceptional preservation of fossils in an upper proterozoic shale. Nature 334:424‐427. 

Casanova, M. T. and M. A. Brock. 2000. How do depth, duration and frequency of flooding influence the establishment of wetland plant communities? Plant Ecology 147:237‐250. 

Chambers, P. A., P. Lacoul, K. J. Murphy, and S. M. Thomaz. 2008. Global diversity of aquatic macrophytes in freshwater. Hydrobiologia 595:9‐26. 

Clavel, J., R. Julliard, and V. Devictor. 2011. Worldwide decline of specialist species: toward a global functional homogenization? Frontiers in Ecology and the Environment 9:222‐228. 

Coops, H. and S. H. Hosper. 2002. Water‐level management as a tool for the restoration of shallow lakes in the Netherlands. Lake and Reservoir Management 18:293‐298. 

Cornwell, W. K. and D. D. Ackerly. 2009. Community assembly and shifts in plant trait distributions across an environmental gradient in coastal California. Ecological Monographs 79:109‐126. 

Degroot, C. J. and C. van Wijck. 1993. The impact of disiccation of a frewater marsh (Garcine Nord, Camargue, France) on sediment water interactions.  

1. The sediment chemistry. Hydrobiologia 252:83‐94. 

 Diaz, S., J. G. Hodgson, K. Thompson, M. Cabido, J. H. 

C. Cornelissen, A. Jalili, G. Montserrat‐Marti, J. P. Grime, F. Zarrinkamar, Y. Asri, S. R. Band, S. Basconcelo, P. Castro‐Diez, G. Funes, B. Hamzehee, M. Khoshnevi, N. Perez‐Harguindeguy, M. C. Perez‐Rontome, F. A. Shirvany, F. Vendramini, S. Yazdani, R. Abbas‐Azimi, A. Bogaard, S. Boustani, M. Charles, M. Dehghan, L. de Torres‐Espuny, V. Falczuk, J. Guerrero‐Campo, A. Hynd, G. Jones, E. Kowsary, F. Kazemi‐Saeed, M. Maestro‐Martinez, A. Romo‐Diez, S. Shaw, B. Siavash, P. Villar‐Salvador, and M. R. Zak. 2004. The plant traits that drive ecosystems: Evidence from three continents. Journal of Vegetation Science 15:295‐304. 

Doledec, S., D. Chessel, and C. Gimaret‐Carpentier. 2000. Niche separation in community analysis: A new method. Ecology 81:2914‐2927. 

Farmer, A. M. and D. H. N. Spence. 1986. The Growth strategies and distribution of isoedtids in Scottish freshwater lochs. Aquatic Botany 26:247‐258. 

Gärdenfors, U., editor. 2010. The 2010 red list of Swedish species. Artsdatabanken, SLU, Upsala. 

Gaston, K. J. and J. I. Spicer. 2001. The relationship between range size and niche breadth: a test using five species of Gammarus (Amphipoda). Global Ecology and Biogeography 10:179‐188. 

Hostrup, O. and G. Wiegleb. 1991. Anatomy of leaves of submerged and emergent forms of Littorella uniflora (L.) Ascherson. Aquatic Botany 39:195‐209. 

Hutchinson, G. E. 1957. Populations studies ‐ animal ecology and demography ‐ concluding remarks. Cold Spring Harbor Symposia on Quantitative Biology 22:415‐427. 

James, C., J. Fisher, V. Russell, S. Collings, and B. Moss. 2005. Nitrate availability and hydrophyte species richness in shallow lakes. Freshwater Biology 50:1049‐1063. 

James, W. F., J. W. Barko, H. L. Eakin, and D. R. Helsel. 2001. Changes in sediment characteristics following drawdown of Big Muskego Lake, Wisconsin. Archiv Fur Hydrobiologie 151:459‐474. 

Jeppesen, E., M. Søndergaard, M. Søndergaard, and K. Christoffersen, editors. 1998. The Structuring Role of Submerged Macrophytes in Lakes. Springer, New York, USA. 

Kålås, J. A., Å. Viken, S. Henriksen, and S. Skjelseth. 2010. The 2010 Norwegian Red List for Species. Norwegian Biodiversity Information Centre. 

Kalff, J. 2002. Limnology: inland water ecosystems. Prentice Hall, Upper Saddle River, USA. 

Karol, K. G., R. M. McCourt, M. T. Cimino, and C. F. Delwiche. 2001. The closest living relatives of land plants. Science 294:2351‐2353. 

27

References 

  

Keddy, P. A. 1992. Assembly and response rules ‐ two goals for predictive community ecology. Journal of Vegetation Science 3:157‐164. 

Kirk, J. T. O. 1994. Light and photosynthesis in aquatic systems. 2nd edition. Cambridge Unirversity Press, Cambridge, UK. 

Koistinen, M. 2010. Stoneworts, Characeae. in P. Rassi, E. Hyvärinen, A. Juslén, and I. Mannerkoski, editors. The 2010 Red List of Finish Species. Ympäristöministeriö & Suomen ympäristökeskus, Helsinki. 

Krause, W. 1997. Charales (Charophyceae).in H. Ettl, G. Gärtner, H. Heynig, and D. Mollenhauer, editors. Süsswasserflora von Mitteleuropa. Gustav Fisher, Stuttgart. 

Lachavanne, J. B., J. Perfetta, and R. Juge. 1992. Influence of water eutrophication on the macrophytic vegetation of Lake Lugago. Aquatic Sciences 54:351‐363. 

Maberly, S. C. and T. V. Madsen. 2002. Freshwater angiosperm carbon concentrating mechanisms: processes and patterns. Functional Plant Biology 29:393‐405. 

Madsen, T. V. and K. Sand‐Jensen. 1991. Photosynthetic carbon assimilation in aquatic macrophytes. Aquatic Botany 41:5‐40. 

Martin‐Closas, C. 2003. The fossil record and evolution of freshwater plants: A review. Geologica Acta 1:315‐338. 

McGill, B. J., B. J. Enquist, E. Weiher, and M. Westoby. 2006. Rebuilding community ecology from functional traits. Trends in Ecology & Evolution 21:178‐185. 

McKinney, M. L. and J. L. Lockwood. 1999. Biotic homogenization: a few winners replacing many losers in the next mass extinction. Trends in Ecology & Evolution 14:450‐453. 

Mjelde, M., S. Hellsten, and F. Ecke. 2013. A water level drawdown index for aquatic macrophytes in Nordic lakes. Hydrobiologia 704:141‐151. 

Møller, C. L. and K. Sand‐Jensen. 2011. High sensitivity of Lobelia dortmanna to sediment oxygen depletion following organic enrichment. New Phytologist 190:320‐331. 

Moss, B. 1998. Ecology of Fresh Waters: Man and Medium, past to future  3edition. Wiley‐Blackwell, Hoboken. 

Nielsen, S. L., E. Gacia, and K. Sand‐Jensen. 1991. Land plants of amphibious Littorella uniflora (L.) Aschers. maintain utilization of CO2 from the sediment. Oecologia 88:258‐262. 

Olsen, S. 1944. Danish Charophyta. Det Kongelige Danske Videnskabernes Selskab, Copenhagen. 

Pedersen, O. and K. Sand‐Jensen. 1993. Water transport in submerged macrophytes. Aquatic Botany 44:385‐406. 

Raun, A. L., J. Borum, and K. Sand‐Jensen. 2010. Influence of sediment organic enrichment and water alkalinity on growth of aquatic isoetid and elodeid plants. Freshwater Biology 55:1891‐1904. 

Raven, P. A., R. F. Evert, and S. E. Eichhorn. 2003. Biology of plants. W. H. Freeman and Company, New York, USA. 

Robe, W. E. and H. Griffiths. 1998. Adaptations for an amphibious life: changes in leaf morphology, growth rate, carbon and nitrogen investment, and reproduction during adjustment to emersion by the freshwater macrophyte Littorella uniflora. New Phytologist 140:9‐23. 

Sand‐Jensen, K., J. Borum, and T. Binzer. 2005. Oxygen stress and reduced growth of Lobelia dortmanna in sandy lake sediments subject to organic enrichment. Freshwater Biology 50:1034‐1048. 

Sand‐Jensen, K., N. L. Pedersen, I. Thorsgaard, B. Moeslund, J. Borum, and K. P. Brodersen. 2008. 100 years of vegetation decline and recovery in Lake Fure, Denmark. Journal of Ecology 96:260‐271. 

Sand‐Jensen, K., T. Riis, O. Vestergaard, and S. E. Larsen. 2000. Macrophyte decline in Danish lakes and streams over the past 100 years. Journal of Ecology 88:1030‐1040. 

Schutten, J., J. Dainty, and A. J. Davy. 2005. Root anchorage and its significance for submerged plants in shallow lakes. Journal of Ecology 93:556‐571. 

Sculthorpe, C. D. 1967. The biology of aquatic vascular plants. Edward Arnold, London, UK. 

Shipley, B. 2010. From plant traits to vegetation structure ‐ Chance and selection in the assembly of ecological communities. Cambridge University Press, Cambridge. 

Shipley, B., D. Vile, and E. Garnier. 2006. From plant traits to plant communities: A statistical mechanistic approach to biodiversity. Science 314:812‐814. 

Silvertown, J., M. E. Dodd, D. J. G. Gowing, and J. O. Mountford. 1999. Hydrologically defined niches reveal a basis for species richness in plant communities. Nature 400:61‐63. 

Simons, J. and E. Nat. 1996. Past and present distribution of stoneworts (Characeae) in the Netherlands. Hydrobiologia 340:127‐135. 

Thompson, K., K. J. Gaston, and S. R. Band. 1999. Range size, dispersal and niche breadth in the herbaceous flora of central England. Journal of Ecology 87:150‐155. 

28

References 

  

Van den Berg, M. S., H. Coops, J. Simons, and A. de Keizer. 1998. Competition between Chara aspera and Potamogeton pectinatus as a function of temperature and light. Aquatic Botany 60:241‐250. 

van der Valk, A. G. 2005. Water‐level fluctuations in North American prairie wetlands. Hydrobiologia 539:171‐188. 

Van Wichelen, J., S. Declerck, K. Muylaert, I. Hoste, V. Geenens, J. Vandekerkhove, E. Michels, N. De Pauw, M. Hoffmann, L. De Meester, and W. Vyverman. 2007. The importance of drawdown and sediment removal for the restoration of the eutrophied shallow Lake Kraenepoel (Belgium). Hydrobiologia 584:291‐303. 

Vestergaard, O. and K. Sand‐Jensen. 2000. Aquatic macrophyte richness in Danish lakes in relation to alkalinity, transparency, and lake area. Canadian Journal of Fisheries and Aquatic Sciences 57:2022‐2031. 

Violle, C., A. Bonis, M. Plantegenest, C. Cudennec, C. Damgaard, B. Marion, D. Le Cæur, and J. B. Bouzillé. 2011. Plant functional traits capture 

species richness variations along a flooding gradient. Oikos 120:389‐398. 

Violle, C. and L. Jiang. 2009. Towards a trait‐based quantification of species niche. Journal of Plant Ecology 2:87‐93. 

Wetzel, R. G. 2001. Limnology: Lake and River Ecosystems. Third edition. Elsevier, San Diego, California. 

Wind, P. and S. Pihl, editors. 2010. The Danish Red List. The National Environmental Research Institute, Aarhus University http://redlist.dmu.dk. 

Wright, I. J., P. B. Reich, M. Westoby, D. D. Ackerly, Z. Baruch, F. Bongers, J. Cavender‐Bares, T. Chapin, J. H. C. Cornelissen, M. Diemer, J. Flexas, E. Garnier, P. K. Groom, J. Gulias, K. Hikosaka, B. B. Lamont, T. Lee, W. Lee, C. Lusk, J. J. Midgley, M. L. Navas, U. Niinemets, J. Oleksyn, N. Osada, H. Poorter, P. Poot, L. Prior, V. I. Pyankov, C. Roumet, S. C. Thomas, M. G. Tjoelker, E. J. Veneklaas, and R. Villar. 2004. The worldwide leaf economics spectrum. Nature 428:821‐827. 

 

 

 

 

29

 

30

31

 

32

Seventy years of changes in the abundance of Danishcharophytes

LARS BAASTRUP-SPOHR, LARS LØNSMANN IVERSEN, JEPPE DAHL-NIELSEN AND

KAJ SAND-JENSEN

Freshwater Biological Laboratory, Biological Institute, University of Copenhagen, Hillerød, Denmark

SUMMARY

1. Charophytes grow attached to soft bottoms in ponds, streams, lakes and estuaries and are highly

threatened throughout Europe according to the national Red Lists. We used Danish studies on fresh-

water charophyte distributions conducted around 1940 and repeated measurements during recent

years to evaluate the historical development of species richness and dominance patterns. We also

tested to what extent historical changes of species abundance in 29 waterbodies were related to

landscape features, water quality and species traits.

2. We found that three species of freshwater charophytes (Chara filiformis, Tolypella intricata and

Nitella gracilis) have apparently disappeared from Denmark while one species (Chara connivens) has

immigrated. National species richness has thus declined from 21 to 19 species.

3. Species abundance based on occurrence in many waterbodies followed a linear rank–log abun-

dance relationship both in the historical and the recent studies. The dominance structure was

stronger today than historically as common species have become relatively more abundant and

uncommon species relatively rarer.

4. Among species traits, perenniality and preference for alkaline waters typical of deep-growing species

in large alkaline lakes, a rare contemporary habitat, were significantly related to the historic species

decline. Species increasing in abundance had wide tolerances to alkalinity and water nitrogen content.

5. Twenty-nine lakes and ponds studied repeatedly showed a significant decline of mean species

richness from 3.4 to 2.4 during the 70 years. A small increase in species richness has taken place

during the recent 15–20 years in several lakes experiencing reduced nutrient loading. However,

many species survive today in relict populations and may find it difficult to recolonise lakes in which

water quality has improved.

6. The historical decline of species richness was significantly related to higher nutrient concentrations,

higher phytoplankton biomass and lower transparency of eutrophied waterbodies. In contrast, the

amount of wetlands and openness of the landscape close to a waterbody did not predict the historical

development perhaps because local processes or long-distance dispersal determine species richness.

7. Considering together the loss of former freshwater habitats and the deterioration of surviving habitats,

we conclude that charophyte occurrence has declined by about 56% in Denmark during the last 70 years.

Species reductions determined from the reinvestigated waterbodies (29%) and the national species list

(10%) are much lower and are less suitable measures of the developmental status of charophytes.

Keywords: charophytes, eutrophication, freshwater stoneworts, habitat deterioration, historical changes

Introduction

Charophytes are 5–200 cm tall macroalgae which are

apparently highly threatened by disappearance from soft-

bottom sediments in lakes, ponds, streams and brackish

waters in Nordic and most other European countries (Bla-

zencic et al., 2006; Siemenska et al., 2006; Langangen,

2007). The abundance of freshwater macrophytes in their

1Correspondence: Lars Baastrup-Spohr, Freshwater Biological Laboratory, Biological Institute, University of Copenhagen, Helsingørsgade 51,

3400 Hillerød, Denmark. E-mail: lbaastrupspohr@bio.ku.dk

© 2013 John Wiley & Sons Ltd 1

Freshwater Biology (2013) doi:10.1111/fwb.12159

33

characteristic environment is significantly related to the

specific ecological traits of the species (Sand-Jensen et al.,

2000). Among charophytes, the largest species appear to

be found in clear lakes while the smaller species can also

be found in turbid waters (Blindow, 1992a; Sand-Jensen

et al., 2000). Because of the rarity, the threatened status

and the markedly different ecological traits of charophyte

species, it is particularly important to explore how their

distribution and relative abundance have developed dur-

ing the previous century characterised by intensive

human impact on aquatic ecosystems. Historical analyses

of their distribution in Switzerland and the Netherlands

have shown the decline of particularly the rare species

during the last 100–200 years (Simons & Nat, 1996; Aud-

erset Joye, Castella & Lachavanne, 2002). In Sweden, a

reduction in charophyte populations has been observed

during the 20th century, probably due to worsened light

conditions caused by eutrophication (Blindow, 1992a).

These changes have not, however, been directly related to

habitat changes or to multiple ecological species traits.

Here, we attempt to fill this gap in our knowledge by

establishing the possible causes behind changes in the

abundance of Danish freshwater charophyte species dur-

ing the last 70 years.

A comprehensive study of the distribution of charo-

phytes in Danish waters is available from the early

1900s with the most extensive data being sampled

around 1940 (Olsen, 1944). Recent investigations (1990–

2011) with most data available for 2009–10 have allowed

us to quantify the development of charophyte vegetation

between 1940 and 2010. The historical analysis is compli-

cated in Denmark by the fact that eutrophication peaked

during the period 1980–95, but subsequently declined

up to 2010 (Jensen et al., 2011). While increasing nutrient

richness and water turbidity from 1900 to 1980–95 may

have led to a reduction of species richness of charo-

phytes, the recent decline of nutrient concentrations and

water turbidity may have increased species richness of

charophytes in those habitats experiencing the greatest

improvement of water quality. Therefore, we expect a

mixed response of species richness and composition in

those 29 lakes and ponds that have been examined

repeatedly during the past 70 years.

Habitat loss will reduce the possibility of dispersal of

charophytes to existing habitats and increase the proba-

bility of extinction of the rare species. Widespread dete-

rioration of freshwater habitats has mainly involved

cultural eutrophication by organic matter, nitrogen and

phosphorus from households, fish farms and agriculture

(Moss, 1998). In Denmark, loading of nitrogen and

phosphorus to aquatic environments increased during

the 20th century with loadings starting to decrease in

the 1970s for phosphorus and 1990s for nitrogen

(Miljøstyrelsen, 1991; Kronvang, 2001), although nutrient

loadings are still markedly above pre-1900 levels (Jensen

et al., 2011). Higher concentrations of nitrogen and phos-

phorus in fresh waters have generally stimulated the

growth and biomass development of phytoplankton and

fast-growing filamentous algae. The resultant increased

shading upon submerged flowering plants and charo-

phytes led either to a decline of their depth penetration

or to total disappearance, depending on the severity of

eutrophication and the type of organism (Sand-Jensen &

Borum, 1991; Middelboe & Markager, 1997). Thus, in

waters experiencing intermediate levels of nutrient

enrichment, tall, robust flowering plants capable of

developing their canopy just below the water surface

may have survived, while competitively inferior bottom-

dwelling mosses and charophytes may have disap-

peared (Middelboe & Markager, 1997; Van Den Berg

et al., 1998a; Sand-Jensen et al., 2008). Several charo-

phytes have a high reproductive potential and, in turn, a

high dispersal capacity via birds and for that reason can

become established in new or recently disturbed lakes with

a restricted coverage of rooted plants (Bonis & Grillas,

2002). We therefore anticipate that the charophyte vege-

tation has changed during the last 70 years towards a

stronger dominance of species tolerant of nutrient-

rich and more perturbed conditions, while species spec-

ialised to oligotrophic, stable conditions should have

declined in abundance because this habitat type has

become much rarer.

Aside from habitat-specific eutrophication, the occur-

rence of aquatic plants has also been shown to be posi-

tively affected by the species richness in well-connected

aquatic habitats because of better chances of colonisation

(Dahlgren & Ehrlen, 2005; Akasaka & Takamura, 2011).

Therefore, we expect that, in the reinvestigated Danish

lakes, loss of charophyte species has been less pro-

nounced in lakes surrounded by many other lakes and

ponds.

Drainage of lowland areas to increase the area of

arable land and eutrophication are virtually universal

and have played a pivotal role in reducing the ecologi-

cal status of freshwater systems across the whole of

Europe (Moss, 1998). We therefore predict that the

same tolerant species have become dominant through-

out Europe and that the species undergoing relative

increases or reductions in abundance in Denmark dur-

ing the last 70 years should have behaved similarly in

other parts of Europe experiencing the same environ-

mental impact.

© 2013 John Wiley & Sons Ltd, Freshwater Biology, doi: 10.1111/fwb.12159

2 L. Baastrup-Spohr et al.

34

Materials and methods

Data collection and taxonomy

The historical distribution of charophyte species was

examined in 114 waterbodies with charophyte vegeta-

tion during the period 1937–42 (Olsen, 1944). The con-

temporary study between 1990 and 2011 included 208

localities with charophyte vegetation examined by the

environmental agencies, university researchers (Nygaard

& Sand-Jensen, 1981; Vestergaard & Sand-Jensen, 2000;

Riis & Sand-Jensen, 2001; Sand-Jensen et al., 2008; Anon-

ymous, 2012) gifted botanists (K. Buchwald, M. Holmen,

B. Moeslund and J. C. Schou, pers. comm. 2010–12) and

recently by ourselves (Dahl-Nielsen, 2011). In addition,

the examination yielded the national species richness in

fresh water in the two periods, including rare species

only found outside the waterbodies under focus. Data

from the old and contemporary studies are referred to

as 1940 and 2010, respectively because these were the

years with most information available.

Charophytes were searched for in lakes, ponds, pits,

streams and ditches. To avoid confounding effects of

salinity, we focused on data from freshwater localities

with salinity lower than 0.5& NaCl. Contemporary

vegetation analyses followed national guidelines (Kris-

tensen et al., 1990) and are regarded as more thorough

in large lakes than old studies because of the recent

use of Scuba diving and sampling along several tran-

sects. In the other waterbodies, thoroughness was prob-

ably the same, but Sigurd Olsen (1944), being an

international expert on charophytes, was perhaps more

skilled at identifying charophyte species than recent

investigators. Altogether there were high numbers of

sites and charophyte species records in both the old

investigation (114 sites, 205 records) and the recent

investigation (170 sites, 295 records) allowing compara-

tive measurements of the presence and abundance of

species.

Charophyte species were identified according to Olsen

(1944) and Blindow et al. (2007). When vegetation analy-

ses were available from several visits, a cumulative spe-

cies list was established. Nomenclature followed

German recommendations (Arbeitsgruppe Characeen

Deutschlands, 2010) in both the old and the recent

study. Chara fragilis was historically the most common

Danish species; it is now named Chara globularis. Histori-

cally, Chara virgata was viewed as a variety of C. fragilis

and its specific distribution is thus not known (Olsen,

1944). Accordingly, we regard C. virgata as being present

in 1937–42, although we are unable to determine its rela-

tive abundance. To be consistent, we regard findings of

C. virgata as C. globularis in the further analysis.

Changing abundance and species traits

We determined the frequency of occurrence of every

species in the two studies as the number of records of a

species divided by the number of sites investigated. The

frequency of occurrence is a measure of how abundant a

given species is. To facilitate comparison among all spe-

cies, some of which are very common while others are

very rare or absent in one of the studies, we transformed

the data logarithmically. We used log (x + 1) to ensure

that species that are absent from one of the two studies

are included in the analysis.

We constructed rank-abundance diagrams of species

in historic and recent studies from Denmark by plotting

log (x + 1) versus falling rank of species and comparing

the negative linear slope (Tokeshi, 1993). To compare the

dominance pattern of species in old and recent studies,

we also calculated Pielou’s evenness index (PEX):

PEX ¼X

pilnpiðlnSÞ�1

where pi is the relative abundance of species 1 to S and

S is species number.

We tested whether changes in the abundance of Dan-

ish species from 1940 to 2010 determined as log (x + 1)

(2010) – log (x + 1) (1940) were correlated with individ-

ual species environmental optimum, tolerance values

and life traits. Environmental optimum and tolerance

values were established as the median and coefficient

of variation (SD/�X; SD is standard deviation, �X is

mean) of total phosphorus, total nitrogen, chlorophyll

a, Secchi transparency and alkalinity in their Danish

growth localities (Anonymous, 2012). The value of each

variable for each site was calculated as time-weighted

mean values preferably for the same year as that in

which the species was recorded. However, in some

instances, when environmental data from the relevant

year were missing data from the years immediately

before or after were used. Lakes with fewer than two

annual records of environmental variables were not

included. Environmental optimum and tolerance values

were established for species recorded in five or more

sites.

Trophic status was also established for 12 species in

the Netherlands from the median values of TP and TN

in their growth localities (Simons & Nat, 1996) to evalu-

ate the importance of independence between trophic

© 2013 John Wiley & Sons Ltd, Freshwater Biology, doi: 10.1111/fwb.12159

Historical changes of charophytes 3

35

rank based on distribution patterns in one country (i.e.

the Netherlands) and abundance based on number of

growth localities in another country (i.e. Denmark).

A series of specific traits that could affect the species

abundance could be obtained from the literature on

Scandinavian charophytes (Langangen, 2007) and sup-

plemented by a Polish investigation in the few cases

where Scandinavian data were missing (Gabka, 2009).

The traits included were shoot height, life form, and

production of oospores and bulbils. The maximum shoot

height (cm) was converted to a numerical scale

(0: 0–20 cm, 1: 21–40 cm and 3: >41 cm) to account for

the variation within and among species. Life form data

were included as (i) obligate annual, (ii) obligate peren-

nial and (iii) both annual and perennial. The intensity of

oospore production was provided by Langangen (2007)

for species with low (0) and high (2) production while

species lacking detailed information were considered

intermediary (1). Some species produce bulbils as a

means of vegetative reproduction and these species were

given the value 1, while the rest were assigned the value

0. The oospore and bulbil production were added

together forming an index of reproductive potential (0–

3).

Site-specific changes

An ANOSIM analysis (Clarke, 1993) was applied to eval-

uate whether the composition of charophyte species in

the Danish waters changed between the historic and the

contemporary investigations. This analysis yields a Glo-

bal-R indicating the separation between groups, and a

significance level can be determined for an unbalanced

number of samples. The analysis is therefore well suited

for our type of data. The analysis was performed on

species presence/absence data in the investigated water-

bodies using the program PRIMER 6 (Clarke & War-

wick, 2001).

A number of the sites visited by Olsen (1944), primar-

ily lakes, were also included in the contemporary study

(called reinvestigated lakes) allowing an evaluation of

the development in species richness and its possible

causes during the last approximately 70 years.

Changes in species richness between the two different

periods were tested with a paired t-test. Also available

for these reinvestigated lakes were several environmen-

tal parameters (total phosphorus, total nitrogen, phyto-

plankton chlorophyll a, Secchi transparency and

alkalinity) collected by the counties and other Danish

environmental agencies during the last 30–40 years

(Anonymous, 2012). Correlation analysis was used to

analyse whether the changes in richness of charophytes

were related to the environmental parameters. In order

to determine whether changes in richness of charophytes

were related to past or present environmental lake con-

ditions, we created simple multiple linear regression

models. The environmental variables were split into pre-

and post-1996 observations, normalised and added into

a PCA. Using relative changes in richness as a response

variable and the primary axis score from the two PCAs

as the explaining variables, the model was reduced by

an F-test using a pre-specified significance level of 5%.

To investigate whether landscape features influenced

the development of the charophyte species richness in

the reinvestigated lakes, we used a GIS-based methodol-

ogy (ArcGIS 10, www.ESRI.com) to derive landscape

variables from a contemporary national landcover map

(AIS landcover map, www. ais.dmu.dk). We selected a

group of features known to affect the viability and dis-

tribution of biodiversity at the landscape level. Firstly,

we determined possible sources of species emigration

within three lag distance to each study lake; these were

expressed as (i) the % coverage of wetlands near a rein-

vestigated lake, (ii) the number of lakes >1 ha and (iii)

the number of ponds and lakes <1 ha. The three lag dis-

tances were chosen to express local landscape-level

effects from the three variables mentioned above. These

distances were chosen in order to cover neighbouring

effects (1 km buffer zone from each study site), long-dis-

tance effects (buffer zones from 5 to 10 km from each

site) and intermediate distance effects (buffer zones from

1 to 5 km from each site). We also determined the dis-

tance to a potentially large dispersal source (distance to

a lake >10 ha), distance to nearest inner coastal zone

and lake surface area, and we determined whether a

locality was interconnected with other lakes since this

might also affect the contemporary charophyte richness.

Plant species depending on wind dispersal (Nathan

et al., 2002) and charophyte dispersal depending on

water birds travelling across open areas might be nega-

tively affected by the amount of forest and thereby the

structure of the landscape. Hence, to express the open-

ness of the surrounding landscape, the amount of forest

was calculated within the three lag distances from each

study lake.

To describe the relationship between landscape vari-

ables and trends in charophyte species richness, we con-

ducted two selective model reduction steps. Firstly, we

fitted four individual logistic regression models of the

four variables described in three lag distances from each

site (amount of forest, amount of wetlands, number of

ponds and number of lakes >1 ha). We used a coding

© 2013 John Wiley & Sons Ltd, Freshwater Biology, doi: 10.1111/fwb.12159

4 L. Baastrup-Spohr et al.

36

(1/0) as response variable, where 1 represents a positive

or stable trend in species diversity, and 0 represents a

decline in diversity. In order to obtain convergence in

the models, only main effects of the three lag distances

of each variable were used (Table 1). We performed

model reduction via a v²-test and removed the least sig-

nificant variable until only significant variables

remained. All resulting P-values were evaluated at a 5%

significance level. Secondly, we created a final logistic

regression model containing significant variables from

the first selection step together with distance to inner

coast, distance to a lake larger than 10 ha, lake surface

area and whether the locality was interconnected with

other lakes (Table 1). Since the viability and current rich-

ness of charophytes are affected by nutrient conditions

within each locality (Blindow, 1992a; Van Den Berg

et al., 1998b), we adjusted for this by adding the primary

axis score from a PCA of the basic environmental vari-

ables within the lakes into the model as a confounding

variable. The final model underwent the same reduction

procedure as the four initial models. All statistical analy-

ses were conducted in R v. 2.14.2 (www.r-projekt.com).

As a subset of the reinvestigated lakes was analysed sev-

eral times during the 70-year period, it was possible to

investigate changes in species richness in the lakes dur-

ing the period of progressive eutrophication from 1940

up to 1975–95 followed by the most recent period of oli-

gotrophication up to 2010.

Results

Loss and gain of species

Three charophyte species have apparently disappeared

from Danish fresh waters and one species has appeared.

Tolypella intricata, Nitella gracilis and the extremely rare

Chara filiformis have most likely gone extinct. Chara

connivens has presumably immigrated to Denmark

between 1940 and 2010. Thus, the historic freshwater

charophyte vegetation included 21 species, while the con-

temporary vegetation includes 19 species.

Common and rare species

The two most common contemporary Danish freshwater

species, Chara globularis and C. vulgaris, were also

the most abundant 70 years ago (Table 2). While

C. globularis has retained its relative abundance, C. vulga-

ris and Nitella flexilis (the third most abundant species)

have increased greatly in relative abundance.

Three species mainly associated with deep oligo-

trophic, alkaline lakes have become less abundant and

are threatened by extinction in Denmark or have already

gone extinct: Chara filiformis, C. intermedia and C. rudis.

Nitella opaca, which used to be relatively widespread in

streams and ditches in the past, is also a rare species

today (Table 2).

Table 2 The abundance (frequency of occurrence, %) of freshwater

charophyte species in the historic (1940) and recent investigations

(2010). Also shown are categorical traits for shoot height, lifer form

and reproductive potential. Recordings with asterisk (*) are species

not present in the investigated lakes, but found in Denmark within

the focal period. N/A indicates no information found in the

literature

Species 1940 2010

Shoot

height Life form

Reproductive

potential

Chara aspera 11.4 16.5 1 3 1

Chara connivens 0.0 2.4 0 3 1

Chara contraria 12.3 10.0 1 3 2

Chara denudata 2.6 1.2 1 N/A 1

Chara filiformis 0.9 0.0 1 2 0

Chara globularis 59.6 56.5 2 3 2

Chara hispida 16.7 5.9 1 3 1

Chara intermedia 11.4 0.0* 2 2 1

Chara polyacantha 1.8 0.0* 2 2 2

Chara rudis 7.0 0.6 0 2 1

Chara tomentosa 7.9 4.1 2 2 0

Chara vulgaris 26.3 32.4 2 3 2

Nitella capillaris 0.0* 1.2 0 1 1

Nitella flexilis 3.5 28.2 2 3 1

Nitella gracilis 0.9 0.0 0 3 1

Nitella mucronata 0.9 0.0* 1 2 0

Nitella opaca 5.3 0.0* 2 2 1

Nitella tranlucens 2.6 4.1 1 3 0

Nitellopsis obtusa 6.1 7.6 2 1 1

Tolypella glomerata 2.6 1.2 0 1 1

Tolypella intricata 0.0* 0.0 1 1 1

Tolypella nidifica 0.0* 1.8 0 1 2

Table 1 Environmental and landscape variables measured in the

reinvestigated lakes. The environmental variables were measured

at the peak of eutrophication between 1980 and 1995 and again

after progressive oligotrophication between 1996 and 2010. Land-

scape variables were calculated based on cartographic material

from 2006

Environmental* Landscape†

Total phosphorus (lg P L�1) % land use (forest, wetland)

Total nitrogen (lg N L�1) Number of ponds (<1 ha) ha�1

Chlorophyll a (lg L�1) Number of lakes (>1 ha) ha�1

Secchi transparency (m) Connection to other lakes (yes/no)

Alkalinity (meq L�1) Distance to nearest lake (km)

Distance to inner coastline (km)

*Time-weighted annual average.†Buffer zones at 0–1 km, 1–5 km, 5–10 km distance from the lake.

© 2013 John Wiley & Sons Ltd, Freshwater Biology, doi: 10.1111/fwb.12159

Historical changes of charophytes 5

37

Species abundance and composition patterns

As anticipated, the relative abundances of species in the

old and recent study were significantly positively corre-

lated (Fig. 1). Negative slopes of species abundance

versus falling rank for all species were nonetheless signifi-

cantly steeper in the recent than the old studies (Fig. 2),

reflecting a contemporary community more characterised

by dominance. This interpretation is supported by Pielou′s

evenness index being higher in the 1940 study (0.79) than

in the 2010 study (0.75).

The ANOSIM analysis allowed investigation of changes

in species composition at the sites in the two periods.

Using the 114 historic and 168 contemporary sites with

known charophyte species composition, the ANOSIM

yielded a Global-R of 0.047 (P = 0.01), showing a signifi-

cant change in species composition between the early and

the recent period. The relatively low Global-R value indi-

cates some degree of overlap between the periods.

It is likely that species preferences and species traits

influence changes in species abundance over the last

70 years. We tested the correlation between the environ-

mental optimum and tolerance values (median and coef-

ficient of variation) of the environmental variables

measured at the sites where each species were present

with the differences in abundance between the old and

the recent study (Table 3). The species optimum and tol-

erance of alkalinity was significantly correlated with

changes of abundance, implying that species preferring

high alkalinity have declined in abundance from 1940 to

2010 and that species increasing in abundance have a

broad alkalinity tolerance. Also, the increasing species

have a wider tolerance of lake nitrogen content, suggest-

ing an ability to thrive under different trophic condi-

tions. The trophic status values from the independent

Fig. 2 Abundance versus species rank (x) of the fifteen most abun-

dant Danish freshwater charophytes in 1940 (open circles) and in

2010 (closed boxes). Abundance was calculated as log (x + 1),

where x is the % species frequency of occurrence. The linear regres-

sion line for the fifteen most abundant species was significantly

steeper in 2010 than 1940 (P < 0.01).

Table 3 The environmental optimum and tolerance values of nine Danish charophyte species (median value/coefficient of variation). The

values are calculated from the environmental variables at all the growth sites of the species.

Species Alkalinity* (meq L�1) TP (lg L�1) TN† (lg L�1) Chlorophyll a (lg L�1) Secchi depth (m) Number of sites

Chara aspera 1.99/0.43 50.7/0.80 1220/0.56 12.1/1.17 1.32/0.73 27

Chara contraria 2.18/0.33 52.7/0.79 1044/0.52 13.1/0.74 2.13/0.56 17

Chara globularis 2.15/0.45 62.1/1.26 1227/0.55 14.1/1.03 1.63/0.67 90

Chara hispida 2.60/0.33 21.1/0.81 951/0.37 8.3/0.81 0.98/0.82 10

Chara tomentosa 2.43/0.24 44.0/0.47 1038/0.33 9.9/0.66 1.38/0.74 7

Chara vulgaris 2.45/0.36 50.8/0.69 1189/0.63 14.6/0.90 1.84/0.66 55

Nitella flexilis 0.93/0.88 43.0/1.45 1014/0.75 12.6/1.11 1.64/0.72 46

Nitella translucens 0.42/1.19 29.2/0.75 897/0.78 8.8/0.94 1.64/0.70 6

Nitellopsis obtusa 2.19/0.26 54.5/0.51 1143/0.44 15.0/0.63 1.27/0.72 13

*Both optimum and tolerance values are significantly correlated with the change of species abundance (Spearman P < 0.05).†The tolerance values are significantly correlated with the change of species abundance (Spearman P < 0.05).

Fig. 1 Relationship between the abundance of charophyte species

in early (1940) and recent Danish studies (2010). The dotted line is

the 1 : 1 line. Abundance was calculated as log (x + 1), where x is

the % species frequency of occurrence.

© 2013 John Wiley & Sons Ltd, Freshwater Biology, doi: 10.1111/fwb.12159

6 L. Baastrup-Spohr et al.

38

Dutch investigation (Simons & Nat, 1996) did not

correlate with the changes in abundance of the Danish

charophyte species (data not shown).

Changes in species abundance could be tested against

the independently collected categorical traits (Table 2).

The analyses showed that obligate perennial species

declined significantly more than either obligate annuals

or species capable of exhibiting both life forms (Kruskal-

Wallis P = 0.004). Shoot size and reproductive potential

did not appear to influence the changes in abundance

(Spearman rank, P > 0.05). Accordingly, it appears that

perennial species and species with preference for highly

alkaline waters are the most vulnerable and subject to a

significant decline, while species tolerant of variable alka-

linity and nitrogen content have increased in abundance.

Species number and composition in revisited Danish

localities

Mean species richness of charophytes in the 29 freshwa-

ter ponds and lakes that were investigated in both 1940

and 2010 was 3.4 in the early investigation and signifi-

cantly lower at 2.4 (paired t-test, P = 0.03) in the recent

investigation. Nine of these waterbodies lost their charo-

phyte vegetation entirely and eight other waterbodies

had reduced species richness in the present investigation

resulting in a decrease in species richness in 58% of the

sites (Fig. 3). In 21% of the sites, no change in richness

was observed while an increase in richness was found

in the remaining 21% sites.

There was a significant linear correlation between sev-

eral of the investigated environmental variables (total

phosphorus (TP), total nitrogen (TN), chlorophyll a, Sec-

chi transparency and alkalinity) and the relative decline

of species richness in the reinvestigated lakes during the

recent decades (1980–95 to 1996–2010) where environ-

mental measurements were available (Table 4). All rela-

tionships suggested that higher nutrient concentrations

increased the loss of original species. The environmental

variables derived from the strongest eutrophication per-

iod (1980–95) had a stronger effect (F = 10.06, P < 0.01)

on the relative changes in richness of charophytes than

the contemporary environmental variables from 1996–

2010 (F = 0.03, P = 0.87).

The four initial models of landscape features indicated

that forest within 5–10 km, number of lakes >1 ha within

5–10 km and number of ponds within 0–1 km from each

site had an effect on the trends in charophyte richness

(Table 5). However, the final model showed that none of

the spatial variables had any significant effects on

changes in diversity (L = 2.77, P = 0.09). This result sug-

gests that the changes in charophyte diversity in Den-

mark are not controlled by any of the spatial variables of

the landscape surrounding the investigated sites.

Recent oligotrophication

Recent oligotrophication of Danish lakes resulting from

the decrease in nutrient loading might have caused an

increase in species richness of charophytes. Thirteen of

the reinvestigated lakes had botanical inventories made

in 1940, 1980–95 and after 2000 making it possible to

search for an effect of the recent oligotrophication (Fig. 4).

Paired tests showed a significant decline in species rich-

ness from 1940 to 1980–95 (Wilcoxon, P = 0.03) while no

differences were found between the other periods for this

Fig. 3 Species richness of charophytes in 29 reinvestigated ponds

and lakes in the past (1940) and recently (2010). The dotted line is

the 1 : 1 line.

Table 4 Correlations between the relative change of species rich-

ness from 1940 to 2010 in the reinvestigated lakes and environmen-

tal variables measured at the peak of eutrophication (1980–95) and

more recently after some nutrient reduction (1996–2010). Positivecorrelation coefficients indicate that lakes with a high value of a

variable in the specific period lost a higher percentage of the origi-

nal species richness

Environmental variable

Correlation

coefficient

(P-value)

Total phosphorus 1980–95 0.55 (0.008)

Total phosphorus 1996–2010 0.47 (0.028)

Total nitrogen 1980–95 0.44 (0.041)

Total nitrogen 1996–2010 0.44 (0.04)

Chlorophyll a 1980–95 0.25 (0.27)

Chlorophyll a 1996–2010 0.33 (0.18)

Secchi depth 1980–95 �0.26 (0.27)

Secchi depth 1996–2010 �0.47 (0.028)

© 2013 John Wiley & Sons Ltd, Freshwater Biology, doi: 10.1111/fwb.12159

Historical changes of charophytes 7

39

subsample. In five of the 13 lakes, however, an increase in

species richness was observed between 1980–95 and after

2000, while only one lake lost species (Fig. 4).

Discussion

Historic changes of charophytes

Charophytes were still widespread and abundant in

European countries up to 100–300 years ago in

resource-limited and physically stable habitats such as

cold, low-light bottom waters of deep oligotrophic lakes

and phosphorus-limited calcareous lakes (Hasslow, 1931;

Iversen, 1929; Olsen, 1944; Nygaard & Sand-Jensen,

1981; Klein, 1993). Thanks to rapid dispersal and coloni-

sation by oospores, some charophyte species were and

still are transiently abundant in newly established ponds

and pits before rooted plants gradually take over and

outcompete them (Olsen, 1944; Crawford, 1977; Wade,

1990). Resource-limited, physically stable habitats or

newly established ephemeral habitats have become

much rarer in the more densely populated and inten-

sively cultivated European landscapes of today and

charophytes have found themselves in more marginal

and threatened positions than before (Agger & Brandt,

1988; Biggs et al., 2005; Munsterhjelm, Henricson &

Sandberg-Kilpi, 2008; Sand-Jensen et al., 2008). In Den-

mark, 85% of the species are proposed to belong to Red

List categories (Baastrup-Spohr et al., 2013) and the per-

centage of species on the Red List is 50–87% across

European countries (Schmidt et al., 1996; Blazencic et al.,

2006; Siemenska et al., 2006). These percentages are

higher than for other aquatic plants species groups that

we know of (G€ardenfors, 2010; K�al�as et al., 2010). The

extremely threatened status of charophytes can be

explained by loss of their habitats in transparent oligo-

trophic lakes at the greatest water depths of submerged

macrophyte growth and inability to compensate for tur-

bid waters by developing tall shoots and to compete

with flowering plants in dense communities in nutrient-

rich habitats (Blindow, 1992b; Van Den Berg et al., 1999).

We document here an overall historic decline in spe-

cies number of freshwater charophytes in Denmark from

21 to 19 species during the last 70 years. Species richness

has also declined in Switzerland (26 to 22), (Auderset

Joye et al., 2002). The modest overall decline of national

species richness sounds much less alarming than the high

Table 5 Effect of landscape features on the trends in charophytes

species richness during the last 70 years. The initial landscape, sig-

nificant variables in the initial models and the final landscape

model are shown. The included variables in the model were calcu-

lated in three lag zones surrounding the lake (0–1 km, 1–5 km and

5–10 km). The variables are forest (% cover), wetlands (% cover),

lakes >1 ha (count), ponds <1 ha (count), distance to a large

(>10 ha) lake (km), distance to inner coastline (km), connection to

other waterbodies (yes/no) and the score of a primary PCA axis on

environmental conditions in the lakes

Model Initial model variables

Significant

variables

Forest Richness trend~ forest

0–1 km + forest 1–5 km

+ forest 5–10k

Forest 5–10k

(L = 4.00,

P < 0.05)

Wetland Richness trend ~ wetland

0–1 km + wetland 1–5 km

+ wetland 5–10k

None

(L = 1.54,

P = 0.21)

Lakes > 1 ha Richness trend ~ lakes 0–1 km

+ lakes 1–5 km + lakes 5–10kLakes 5–10k(L = 8.19,

P < 0.01)

Ponds Richness trend ~ ponds 0–1 km

+ ponds 1–5 km + ponds 5–10k

Ponds 0–1 km

(L = 12.32,

P < 0.001)

Final Richness trend ~ PCA1 + area_lake

+ lakes 5–10 km + ponds 0–1 km

+ forest 5–10 km + distance to

coast + distance to lake >10 ha

+ lake_connection

None

(L = 2.77,

P = 0.09)

Fig. 4 Species richness of charophytes in thirteen lakes measured

three times between 1940 and 2010. Five lakes have experienced a

recent increase of species richness (upper panel), whereas no recent

changes occurred in the other lakes aside from one lake experiencing

a minor richness reduction (lower panel). Overall, there was a signifi-

cant decline of species richness from 1940 to 1980–95 (Wilcoxon test

P = 0.02), but no significant change between either 1940 or 1980–95

and the contemporary investigation (P = 0.24 and P = 0.14).

© 2013 John Wiley & Sons Ltd, Freshwater Biology, doi: 10.1111/fwb.12159

8 L. Baastrup-Spohr et al.

40

proportions of species on the national Red Lists because

most species are still present within the countries though

in diminishing populations and at fewer localities. In that

respect, changes in species number in the same localities

offer a more precise quantitative descriptor of the devel-

opment. In 29 of such Danish localities, the mean species

number of charophytes per locality declined from 3.4 to

2.4 (29%) over the past 70 years due to habitat deteriora-

tion. Still, this comparison does not include former charo-

phyte localities, mostly small ponds and pits, which have

been lost entirely by being drained or filled in. Fifty-four

such localities contained, on average, 1.43 species in

Olsen’s (1944) historical study. If we assume that lost

ponds and pits amounted to twice the number of

retained localities (Sand-Jensen et al., 2000; Fog, 2011),

represented by the aforementioned 29 localities with a

mean decline of species richness from 3.4 to 2.4, then the

average historical decline of charophyte occurrences in

Denmark as a combined result of habitat deterioration

and habitat loss would amount to 56%. In comparison,

Danish species richness declined by only 10% and is a

much less sensitive measure of the overall status of

charophytes because one growth site for a species con-

tributes with the same score as numerous growth sites.

Our comparisons among localities document a signifi-

cant correlation between the decline of species richness

and the impoverished water quality due to eutrophica-

tion (i.e. higher nutrient concentrations, phytoplankton

biomass and water turbidity) in accordance with numer-

ous former investigations (Jupp & Spence, 1977; Lachav-

anne, Perfetta & Juge, 1992; Sand-Jensen et al., 2000;

James et al., 2005). In 13 lakes investigated three times

during the 70 years, we also observed an increase in

species richness in five of the lakes after nutrient loading

has been reduced during the last 15 years. Although

this overall increase of charophyte richness was not

statistically significant, it has been quite distinct in lakes

experiencing a great improvement of water quality (Van

Den Berg et al., 1999; Jeppesen et al., 2005; Sand-Jensen

et al., 2008). These observations stress that the charo-

phyte vegetation can recover, although it is questionable

whether full recovery to the pre-1940 level of richness is

possible given that some species have gone nationally

extinct, that several species only survive in relict popula-

tions in a few localities, and that it is doubtful whether

lakes fully return to their former nutrient status.

Processes behind charophyte losses

It is not the eutrophication of waterbodies alone, but a

combination of deterioration of waterbodies (i.e. by

eutrophication and physical disturbance) and total loss

of suitable waterbodies that has led to decline of the

charophyte vegetation in European countries. The loss

or decline of 38 populations of Chara tomentosa and the

stable occurrence or positive development of only seven

populations in the low-saline brackish waters of SE Fin-

land between the 1890s and 2003 illustrate the harmful

human impact (Munsterhjelm et al., 2008). In 27 of the

Finnish localities, the charophyte decline could be linked

to either eutrophication (19 cases of stimulated growth

of filamentous algae and nine cases of higher input of

nutrient-rich drainage water), higher physical impact

(10–11 cases of shore activity, dredging, sediment spread

and boat traffic) or a combination of the two impacts

(i.e. 18 cases of higher water turbidity from phytoplank-

ton and sediment particles). The general loss of water-

bodies with charophyte vegetation can be illustrated by

the vast decline of ponds in Great Britain from 1.2 mil-

lion in 1880 to only 0.3 million in 1990 (Biggs et al.,

2005), and the same 70–80% decline of ponds and shal-

low lakes has taken place in Denmark from 1900 to 1990

(Kristensen, Reenberg & Pena, 2009; Fog, 2011).

We initially proposed, but were unable to confirm,

the existence of significant relationships between the

richness of charophytes in focal waterbodies and fea-

tures of the landscape surrounding the lakes. These

findings suggest that species richness of the individual

waterbody is primarily dependent on its water quality

and that dispersal from waterbodies located close by

is not a critical factor for the occurrence because either

long-distance dispersal or local processes are sufficient

to sustain species richness.

Like terrestrial mosses, lycopods and ferns, charophytes

have ended up in more marginal and resource-poor habi-

tats than the more recently evolved flowering plants

(Bates, 1998). Charophytes and mosses can be expected to

be restricted by the lack of roots, solid stems and vascular

tissue for efficient nutrient uptake and translocation while

lycopods and ferns can be expected to be restricted by the

lack of secondary growth and well-protected seeds com-

pared with flowering plants. Although these functional

shortcomings have excluded charophytes, mosses, lyco-

pods and ferns from resource-rich, unperturbed habitats,

their long-term evolutionary histories have adapted them

to the marginal, resource-poor habitats through their

exceedingly low growth rates, high persistence and capac-

ity for long-distance dispersal of spores between widely

scattered habitats (Hedderson, 1992; Longton, 1992).

Charophytes and aquatic mosses have systematically

lower growth and photosynthetic rates than aquatic

flowering plants (Kautsky, 1988; Nielsen & Sand-Jensen,

© 2013 John Wiley & Sons Ltd, Freshwater Biology, doi: 10.1111/fwb.12159

Historical changes of charophytes 9

41

1989; Riis & Sand-Jensen, 1997; Dahl-Nielsen, 2011), and

terrestrial mosses, likewise, photosynthesise and grow

much slower than flowering plants (Martin & Adamson,

2001; Glime, 2007; Sand-Jensen et al., 2010).

Abundance patterns and coupling of abundance to species

traits

The study confirmed our initial prediction that common

charophyte species have become relatively more abun-

dant and rare species relatively less abundant in Den-

mark during the historic loss and deterioration of

suitable charophyte habitats. The same development has

taken place in the Netherlands and Switzerland (Simons

& Nat, 1996; Auderset Joye et al., 2002). Our analysis

showed that rank-abundance diagrams became signifi-

cantly steeper and indices of evenness smaller in all

three countries (data not shown). A parallel develop-

ment has been observed at the local scale among sub-

merged macrophytes upon eutrophication of Lake Fure

(Sand-Jensen et al., 2008). Terrestrial grassland communi-

ties likewise respond to increasing soil richness by

developing stronger dominance structure among species

in gradually increasing competition for light in denser

communities (Stevens & Carson, 1999).

It is remarkable that Chara vulgaris and C. globularis

are among the two most abundant species in Denmark,

the Netherlands, the Czech Republic, Switzerland and

the Balkans, and that Nitella flexilis is the third most

abundant species in the three first-mentioned countries

(Simons & Nat, 1996; Auderset Joye et al., 2002; Blazen-

cic et al., 2006; Caisova & Gabka, 2009). These three spe-

cies are located as numbers 15, 14 and 12 according to

increasing trophic rank among 16 species in the

Netherlands and are there regarded as being among the

most eutrophic species (Simons & Nat, 1996).

The three species lost from Danish waters were

already rare in the 1940s, with only 1–6 recorded sites

(Olsen, 1944). While both Nitella gracilis and Chara filifor-

mis can be perennial species from more stable habitats,

Tolypella intricata is a spring annual characterised by

rapid appearance in ephemeral habitats. Tolypella intri-

cata was thus thought to be extinct in Sweden, but

spring sampling in former localities revealed the pres-

ence of the species in a few localities (G€ardenfors, 2010).

This recent finding suggests that the species could still

be present in Denmark, although the heavily drained

Danish landscape with very few ephemeral waterbodies

makes this expectation doubtful.

The relationship between the historical changes of

abundance and species traits in Denmark revealed a

decline of perennial species and species with preference

for highly alkaline waters. These species primarily grow,

often to a large depth, in transparent large alkaline

lakes, a rare habitat in the contemporary Danish land-

scape (Bjerring et al., 2011). Species with broader niches

capable of growing in lakes of varying alkalinity and

nitrogen content and, thus, under more perturbed condi-

tions have better been able to persist.

The observed pattern of success of generalist species

and the decline of specialists is consistent with investi-

gations on the response of other types of organisms to

habitat homogenisation (Rooney et al., 2004; Devictor

et al., 2008). Specialists are defined as having a very

high fitness under specific conditions while generalists

have a lower fitness, but can thrive under a wider

range of conditions (Clavel, Julliard & Devictor, 2011).

When conditions change, specialists lose their fitness

advantage and disappear while the more versatile gen-

eralists survive or even thrive. We suggest that the loss

of waterbodies and increases in eutrophication across

Europe have caused the observed changes in charo-

phyte community composition and dominance patterns

all over the continent.

In conclusion, habitat loss and habitat deterioration

by eutrophication have reduced the total occurrence of

charophytes by about 56% in Denmark while mean

species richness in reinvestigated waterbodies declined

by 29% and national species richness by only 10%.

Charophytes are critically threatened with 82% of the

species on the Danish Red List. The dominance struc-

ture has changed such that common species have

become more abundant relative to rare species. Species

with wide tolerances to both alkalinity and nitrogen

content have increased supporting the hypothesis of

generalists being better at surviving in disturbed land-

scapes. Nutrient reduction in several waterbodies dur-

ing the last 15 years has led to increased species

richness, highlighting that historical development is in

part reversible, although several species may be unable

to recover because they survive only in small relict

populations.

Acknowledgments

This work was funded by the Center for Lake Restora-

tion (CLEAR), a Villum Kann Rasmussen Center of

Excellence awarded to KSJ, and by a PhD grant from the

Faculty of Science, University of Copenhagen to LBS.

We are truly grateful to Irmgard Blindow for valuable

comments on an earlier version of this manuscript.

© 2013 John Wiley & Sons Ltd, Freshwater Biology, doi: 10.1111/fwb.12159

10 L. Baastrup-Spohr et al.

42

References

Agger P. & Brandt J. (1988) Dynamics of small biotopes in

Danish agricultural landscapes. Landscape Ecology, 1, 227–

240.

Akasaka M. & Takamura N. (2011) The relative importance

of dispersal and the local environment for species rich-

ness in two aquatic plant growth forms. Oikos, 120, 38–46.

Anonymous (2012) Danish environmental web portal.

Accessed april 2011.

Auderset Joye D., Castella E. & Lachavanne J.B. (2002)

Occurrence of Characeae in Switzerland over the last two

centuries (1800–2000). Aquatic Botany, 72, 369–385.

Baastrup-Spohr L., Dahl-Nielsen J. & Sand-Jensen K. (2013)

Historical changes in abundance of characeans in Denmark

(in Danish). Urt, 37, 66–70.

Bates J.W. (1998) Is ‘life-form’ a useful concept in bryophyte

ecology? Oikos, 82, 223–237.

Biggs J., Williams P., Whitfield M., Nicolet P. & Weatherby

A. (2005) 15 years of pond assessment in Britain: results

and lessons learned from the work of Pond Conservation.

Aquatic Conservation-Marine and Freshwater Ecosystems, 15,

693–714.

Bjerring R., Johansson L.S., Søndergaard M., Kjeldgaard A.,

Sortkjær L., Windolf J. et al. (2011) Lakes 2010 – NOVANA

(in Danish). Danish Center for Environment and Energy

(DCE), Aarhus University.

Blazencic J., Stevanovic B., Blazencic Z. & Stevanovic V.

(2006) Red data list of charophytes in the Balkans. Biodi-

versity and Conservation, 15, 3445–3457.

Blindow I. (1992a) Decline of charophytes during eutrophi-

cation – comparison with angiosperms. Freshwater Biology,

28, 9–14.

Blindow I. (1992b) Long- and short-term dynamics of sub-

merged macrophytes in two shallow eutrophic lakes.

Freshwater Biology, 28, 15–27.

Blindow I., Krause W., Ljungstrand E. & Koistinen M.

(2007) Best€amningsnyckel f€or kransalger i Sverige. Svensk

Botanisk Tidskrift, 101, 165–220.

Bonis A. & Grillas P. (2002) Deposition, germination and

spatio-temporal patterns of charophyte propagule banks:

a review. Aquatic Botany, 72, 235–248.

Caisova L. & Gabka M. (2009) Charophytes (Characeae,

Charophyta) in the Czech Republic: taxonomy, autecol-

ogy and distribution. Fottea, 9, 1–43.

Clarke K.R. (1993) Nonparametric multivariate analyses of

changes in community structure. Australian Journal of

Ecology, 18, 117–143.

Clarke K.R. & Warwick R.M. (2001) Changes in Marine Com-

munities: An Approach to Statistical Analysis and Interpreta-

tion. PRIMER-E, Plymouth.

Clavel J., Julliard R. & Devictor V. (2011) Worldwide

decline of specialist species: toward a global functional

homogenization? Frontiers in Ecology and the Environment,

9, 222–228.

Crawford S.A. (1977) Chemical, physical and biological

changes associated with Chara succession in farm ponds.

Hydrobiologia, 55, 209–217.

Dahlgren J.P. & Ehrlen J. (2005) Distribution patterns of

vascular plants in lakes – the role of metapopulation

dynamics. Ecography, 28, 49–58.

Dahl-Nielsen J. (2011) Characeans: Distribution and Ecology

(in Danish). Master Thesis, University of Copenhagen,

Hillerød.

Devictor V., Julliard R., Clavel J., Jiguet F., Lee A. & Couvet

D. (2008) Functional biotic homogenization of bird com-

munities in disturbed landscapes. Global Ecology and Bio-

geography, 17, 252–261.

Fog K. (2011) The Danish amphibians. In: Danish Nature

2010 -About the Loss of Biological Richness (in Danish). (Ed.

H. Meltofte), pp. 49–54. Det Grønne Kontaktudvalg.

Gabka M. (2009) Charophytes of the Wielkopolska Region (NW

Poland): Distribution, Taxonomy and Autecology. Bogucki

Wydawnictwo Naukowe, Poznan.

G€ardenfors U. (2010) The 2010 Red List of Swedish Species.

Artsdatabanken, SLU, Uppsala.

Glime J.M. (2007) Productivity. In: Bryophyte Ecology -Physio-

logical Ecology (Ed. J.M. Glime), pp. 1–21. Available at:

http://www.bryoecol.mtu.edu/.

Hasslow O.J. (1931) Charophytes of Sweden (in Swedish).

Botaniska Notiser, 63–136.

Hedderson T.A. (1992) Rarity at range limits - dispersal

capacity and habitat relationships of extraneous moss

species in a boreal Canadian national-park. Biological Con-

servation, 59, 113–120.

Iversen J. (1929) Studien €uber pH-Verh€altnisse d€anischer

Gew€asser und ihren Einfluss auf die Hydrophyten-

Vegetation. Botanisk Tidskrift, 40, 277–333.

James C., Fisher J., Russell V., Collings S. & Moss B. (2005)

Nitrate availability and hydrophyte species richness in

shallow lakes. Freshwater Biology, 50, 1049–1063.

Jensen P.N., Boutrup S., Svendsen L.M., Grant R., Windolf

J., Bjerring R. et al. (2011) Aquatic Environment and Nat-

ure 2010 (in Danish). Danish Centre for Environment and

Energy, Aarhus, Denmark.

Jeppesen E., Søndergaard M., Jensen J.P., Havens K.E., Anne-

ville O., Carvalho L. et al. (2005) Lake responses to reduced

nutrient loading - an analysis of contemporary long-term

data from 35 case studies. Freshwater Biology, 50, 1747–1771.

Jupp B.P. & Spence D.H.N. (1977) Limitations on macro-

phytes in a eutrophic lake, Loch Leven 1. Effects of phy-

toplankton. Journal of Ecology, 65, 175–186.

K�al�as J.A., Viken �A., Henriksen S. & Skjelseth S. (2010) The

2010 Norwegian Red List for Species. Norwegian Biodiver-

sity Information Centre, Trondheim, Norway.

Kautsky L. (1988) Life strategies of aquatic soft bottom mac-

rophytes. Oikos, 53, 126–135.

Klein T. (1993) Impact on lake development of changed

agricultural watershed exploitation during the last 3 cen-

turies. Hydrobiologia, 251, 297–308.

© 2013 John Wiley & Sons Ltd, Freshwater Biology, doi: 10.1111/fwb.12159

Historical changes of charophytes 11

43

Kristensen P., Søndergaard M., Jeppesen E., Mortensen E. &

Rebsdorf A. (1990) Sampling and Methods of Analysis in

Lakes. Monitoring Program (in Danish). National Envi-

ronmental Institute, Silkeborg, Denmark.

Kristensen S.B.P., Reenberg A. & Pena J.J.D. (2009) Exploring

local rural landscape changes in Denmark: a human-envi-

ronmental timeline perspective. Geografisk Tidsskrift-Danish

Journal of Geography, 109, 47–67.

Kronvang B. (2001) Phosphorus in Soil and Water: Develop-

ment, status and perspectives (in Danish). National Environ-

mental Research Institue, Silkeborg.

Lachavanne J.B., Perfetta J. & Juge R. (1992) Influence of

water eutrophication on the macrophytic vegetation of

Lake Lugago. Aquatic Sciences, 54, 351–363.

Langangen A. (2007) Charophytes of the Nordic countries.

Saeculum ANS, Oslo.

Longton R.E. (1992) Reproduction and rarity in British

mosses. Biological Conservation, 59, 89–98.

Martin C.E. & Adamson V.J. (2001) Photosynthetic capacity

of mosses relative to vascular plants. Journal of Bryology,

23, 319–323.

Middelboe A.L. & Markager S. (1997) Depth limits and

minimum light requirements of freshwater macrophytes.

Freshwater Biology, 37, 553–568.

Miljøstyrelsen (Ed.) (1991) Aquatic environment -90 (in

Danish). In: Statement from Miljøstyrelsen. Miljøstyrelsen,

Copenhagen.

Moss B. (1998) Ecology of Fresh Waters: Man and Medium, past

to future. Wiley-Blackwell, Hoboken.

Munsterhjelm R., Henricson C. & Sandberg-Kilpi E. (2008)

The decline of a charophyte – occurrence dynamics of

Chara tomentosa L. at the southern coast of Finland. Memo-

randa- Societas pro Fauna et Flora Fennica, 84, 56–80.

Nathan R., Katul G.G., Horn H.S., Thomas S.M., Oren R.,

Avissar R. et al. (2002) Mechanisms of long-distance dis-

persal of seeds by wind. Nature, 418, 409–413.

Nielsen S.L. & Sand-Jensen K. (1989) Regulation of photo-

synthetic rates of submerged rooted macrophytes. Oecolo-

gia, 81, 364–368.

Nygaard G. & Sand-Jensen K. (1981) Light climate and

metabolism of Nitella flexilis (L) Ag in the bottom waters

of oligotrophic Lake Grane Langsø, Denmark. Internatio-

nale Revue Der gesamten Hydrobiologie, 66, 685–699.

Olsen S. (1944) Danish Charophyta. Det Kongelige Danske

Videnskabernes Selskab, Copenhagen.

Riis T. & Sand-Jensen K. (1997) Growth reconstruction and

photosynthesis of aquatic mosses: Influence of light, tem-

perature and carbon dioxide at depth. Journal of Ecology,

85, 359–372.

Riis T. & Sand-Jensen K. (2001) Historical changes in spe-

cies composition and richness accompanying perturbation

and eutrophication of Danish lowland streams over

100 years. Freshwater Biology, 46, 269–280.

Rooney T.P., Wiegmann S.M., Rogers D.A. & Waller D.M.

(2004) Biotic impoverishment and homogenization in un-

fragmented forest understory communities. Conservation

Biology, 18, 787–798.

Sand-Jensen K., Baastrup-Spohr L., Møller C.L., Winkel A.,

Brodersen K.P., Borum J. et al. (2010) Plant distribution

patterns and adaptation in a limestone quarry on €Oland.

Svensk Botanisk Tidsskrift, 104, 23–31.

Sand-Jensen K. & Borum J. (1991) Interactions among phy-

toplankton, periphyton, and macrophytes in temperate

freshwaters and estuaries. Aquatic Botany, 41, 137–175.

Sand-Jensen K., Pedersen N.L., Thorsgaard I., Moeslund B.,

Borum J. & Brodersen K.P. (2008) 100 years of vegetation

decline and recovery in Lake Fure, Denmark. Journal of

Ecology, 96, 260–271.

Sand-Jensen K., Riis T., Vestergaard O. & Larsen S.E. (2000)

Macrophyte decline in Danish lakes and streams over the

past 100 years. Journal of Ecology, 88, 1030–1040.

Schmidt D., Van der Weyer K., Krause W., Kies L., Garniel

A. et al. (1996) Rote Liste der Armleuchteralgen (Charophy-

ceae) Deutschlands, Schriftenreihe f€ur Vegetationskunde.

Siemenska J., Bak M., Dziedzic J., Gabka M., Gregorowicz P.,

Mrozinska T. et al. (2006) Red list of the algae in Poland. In:

Red list of plants and fungi in Poland (Eds Zarzycki K.,

Wojewoda W. & Szelag Z.W.), pp. 37–52. Szafer Institue of

Botany, Polish Academy of Sciences, Krakow.

Simons J. & Nat E. (1996) Past and present distribution of

stoneworts (Characeae) in the Netherlands. Hydrobiologia,

340, 127–135.

Stevens M.H.H. & Carson W.P. (1999) Plant density deter-

mines species richness along an experimental fertility gra-

dient. Ecology, 80, 455–465.

Tokeshi M. (1993) Species abundance patterns and commu-

nity structure. Advances in Ecological Research, 24, 111–186.

Van Den Berg M.S., Coops H., Simons J. & De Keizer A.

(1998a) Competition between Chara aspera and Potamog-

eton pectinatus as a function of temperature and light.

Aquatic Botany, 60, 241–250.

Van Den Berg M.S., Scheffer M., Coops H. & Simons J.

(1998b) The role of characean algae in the management of

eutrophic shallow lakes. Journal of Phycology, 34, 750–756.

Van Den Berg M.S., Scheffer M., Van Nes E. & Coops H.

(1999) Dynamics and stability of Chara sp. and Potamog-

eton pectinatus in a shallow lake changing in eutrophica-

tion level. Hydrobiologia, 408, 335–342.

Vestergaard O. & Sand-Jensen K. (2000) Alkalinity and tro-

phic state regulate aquatic plant distribution in Danish

lakes. Aquatic Botany, 67, 85–107.

Wade P.M. (1990) The colonisation of disturbed freshwater

habitats by Characea. Folia Geobotanica and Phytotaxonomica,

25, 275–278.

(Manuscript accepted 2 April 2013)

© 2013 John Wiley & Sons Ltd, Freshwater Biology, doi: 10.1111/fwb.12159

12 L. Baastrup-Spohr et al.

44

45

 

46

Niche specialization and traits of charophytes in Scandinavia  

  

Niche specialization and functional traits regulate abundance of Scandinavian charophytes Lars Baastrup-Spohr*, Lars Lønsmann Iversen, Jens Borum and Kaj Sand-Jensen Freshwater Biological Laboratory, Biological Institute, University of Copenhagen, Universitets parken 4, 2100 Kbh. Ø.

Summary 1. Charophytes are slow-growing macroalgae that live attached to soft sediments in lakes, ponds, streams and brackish waters. Given the historic loss and eutrophication of their habitats and their competitive inferiority to tall rooted plants, we predict that charophytes are among the most threatened phototrophs. We also predict that charophyte species with generalist niches and functional traits such as high salinity tolerance and large height should be best able to survive due to more growth habitats and stronger competitive capability. 2. We tested these predictions by examining the distribution of charophytes in Scandinavia and the relationship between species abundance and species traits. Abundance values were assigned to each species based on their status on the national Red List. Niche specialization was evaluated by the Outlying Mean Index (OMI) using a large Danish dataset on charophyte distribution and environmental variables. Species functional traits were derived from the literature. 3. The results supported our predictions showing that 50-87% of the species were on the national Red Lists in the four Scandinavian countries. These proportions are much greater than of aquatic (30-35%) and terrestrial plants (18-28%). Species abundance of charophytes in Scandinavia decreased significantly with niche specialization in separate analyses of marine and freshwater sites. Four functional traits: shoot height, salinity tolerance, bulbil production and ability to express both perennial and annual life form were significantly positively related to Scandinavian abundance. 4. We conclude that charophytes are generally rare species restricted by habitat loss and competition with fast-growing phytoplankton, macroalgae and tall vascular plants at eutrophic sites. The few charophyte species maintaining a high abundance in Scandinavia are generalists tolerant of a wide range of conditions, while specialists are often short, have a restricted life form and are rare in the disturbed contemporary landscape. Keywords: Charophyte, niche specialization, functional traits, Red List, species abundance

Introduction

Charophytes belong to an ancient group of macroscopic, slow-growing green algae that live attached by hair-like rhizoids to soft sediments in lakes, ponds, streams and brackish waters (Krause, 1997). Some species are widely distributed from tropical to temperate and subarctic waters across the World, others are distributed in temperate or tropical regions and yet others are confined to narrower regions such as Scandinavia (Langangen, 2007). Most charophytes grow in

oligo- and mesotrophic waters, while they are often outcompeted by phytoplankton, blanketing filamentous algae and tall rooted angiosperms in eutrophic waters (Blindow, 1992, Sand-Jensen, Pedersen, Thorsgaard et al., 2008). We predict that charophytes should include particularly high proportions of threatened species because widespread eutrophication has increased shading from all three mentioned groups of phototrophic competitors (Sand-Jensen et al., 2008).

47

Niche specialization and traits of charophytes in Scandinavia  

  

However, some charophytes survive thanks to their ability to rapidly colonize and establish temporary dominance in newly constructed or perturbed waters because of high production of oospores that are efficiently dispersed by water birds (Crawford, 1977, Bonis & Grillas, 2002, Brochet et al., 2010). Here we examine both the status of charophytes on national Red Lists in Scandinavia and test how variability in species niche specificity and species traits such as shoot height, reproductive investment and life form influence their abundance in Scandinavia.

Charophytes have experienced a dramatic decline in abundance during the last 150 years in all countries with dense human populations and intensive agriculture (Simons & Nat, 1996, Blindow, 2000, Auderset Joye, Castella & Lachavanne, 2002, Baastrup-Spohr et al., 2013b). Widespread cultural eutrophication has been the most important environmental deterioration of charophyte habitats leading to restriction of depth distribution or total eradication of charophyte species in numerous water bodies (Blindow, 1992, 2000, Baastrup-Spohr et al., 2013b). Also mechanical disturbance by dredging, construction work, motor boat traffic and reed cutting can disturb sediment stability, increase turbulence and turbidity and, thereby, eradicate charophytes (Torn et al., 2010). A third environmental deterioration, mainly operating during the 1950-1980s, involved acidification of low-alkaline waters because of high sulfur deposition from the atmosphere leading to replacement of charophytes by acid-tolerant mosses and filamentous green algae (Riis & Sand-Jensen, 1998). A final important threat has been disappearance of numerous small water bodies and establishment of only a few new ones because of intensified agricultural exploitation of the landscape (Biggs et al., 2005). As a consequence, many charophytes have become rare, threatened by extinction and placed on the IUCN Red List in European countries, while the remaining common species have not declined to the same extent (Blazencic et al., 2006, Gärdenfors, 2010,

Baastrup-Spohr et al., 2013b). Although charophytes have been subject to widespread and severe historical decline (Simons & Nat, 1996, Blindow, 2000, Baastrup-Spohr et al., 2013b) there has been no concerted effort to document their general status and evaluate the possible determinants of species abundance. Here we provide this information by determining the proportions of nationally Red-listed charophyte species. We determine the similarity in species composition and abundance between the Scandinavian countries. High similarity would support the notion that it is possible to identify mechanisms that determine whether species are common or rare.

High habitat specificity should by definition restrict species to a few special, but very suitable habitats and they are suggested to remain rare overall because the majority of habitats would be unsuited for them (Brown, 1984). This concept of generalists and specialists is thus based on the trade-off between the capacity to exploit or tolerate a range of conditions (generalists) and performing better under specific conditions (Futuyma & Moreno, 1988). Furthermore, in disturbed contemporary landscapes a similar pattern in the change of species composition has been observed for several species groups; i.e. a decline of specialist and increase of generalist species (Clavel, Julliard & Devictor, 2011).

Therefore, we predict that the degree of specialization of species with respect to physico-chemical and biological habitat conditions should be negatively related to the abundance of charophytes in Scandinavia.

The response of species to environmental gradients are mediated through functional traits (Keddy, 1992). Conceptually, environmental conditions can be viewed as a set of filters only letting through those species with the suitable traits for a given filter (Keddy, 1992). As species often face multiple abiotic and biotic gradients several traits can be involved in determining community composition (Shipley, 2010).

48

Niche specialization and traits of charophytes in Scandinavia  

  

For flowering aquatic plants, several traits have been shown to be significantly related to abundance and, in particular, plant height was related to abundance in both lakes and streams (Sand-Jensen et al., 2000). We thus hypothesize the taller charophytes should be more abundant than shorter species due to stronger competitive ability for light and space of the taller species.

Tolerance of salinity in charophyte species is related to the physiological ability of the species to regulate cell turgor pressure (Winter, Soulie-Märsche & Kirst, 1996) and experiments with several species have shown different salinity tolerances (Winter et al., 1996, Blindow et al., 2003). The observed tolerances correspond well with the distribution of the species within different salinity zones of the Baltic Sea (Blindow, 2000). Because of the close proximity of brackish and freshwater habitats in the Scandinavian region and the large quantity of brackish water habitats we expect that increasing salinity tolerance should positively influence regional abundance.

Charophyte propagules are dispersed with the large number of migratory waterfowl (Bonis & Grillas, 2002, Brochet et al., 2010), and we predict that production of oospores and vegetative propagules by charophyte species should be positively related to the likelihood of dispersal and thereby to the regional abundance. Oospore formation requires the involvement of both male antheridia and female oogonia and these are present on the same individual shoot in most charophyte species (monoecious), while fewer species (dioecious) have separate male and female shoots (Langangen, 2007). When species become rare, monoecious species can probably better maintain oospore formation, while dioecious species can ensure outcrossing and better maintain genetic variability and life vitality associated with small populations. Therefore, we have no a priori expectation as to whether common or

rare species should have a larger proportion of monoecious or dioecious species.

Other biological traits may also influence charophyte growth and survival in their specific habitats and, thereby, their national rarity or commonness. Certain varieties within species may differ by being either annual or perennial, while other species are reported as being permanently annual or perennial. For example, a summer annual form should be selectively advantageous in shallow pond and lake sediments disturbed by winter ice-scouring, while perennial forms should be advantageous in low-light and physically stabile deep waters where long time is required but also available for population development. We, therefore, predict that species housing both annual and perennial life forms can occupy more habitats than species housing only one of the two life forms.

Thus we set out to test how niche specialization and functional traits may account for the abundance of Scandinavian charophytes.

Materials and methods

Study area

Scandinavia in this context consists of Denmark, Sweden, Norway and Finland (Fig. 1). Denmark (DK) is the southernmost and the smallest country (43,095 km2) with the fewest freshwater bodies. Nonetheless, because of calcareous soils and high-alkaline freshwaters suitable for a high variety of Chara species, the country had a very rich charophyte vegetation 70 years ago (Olsen, 1944) before widespread eutrophication and drainage of approximately two-thirds of all water bodies led to profound impoverishment (Baastrup-Spohr et al., 2013b). Sweden (SW) is the largest (449,964 km2) and most variable country with respect to geology, terrain and water types and had the richest charophyte vegetation in Scandinavia 70 years ago (Olsen, 1944).

49

Niche specialization of charophytes  

________________________________________________________________________________ For correspondence: E-mail: lbaastruspohr@bio.ku.dk Freshwater Biological Laboratory, Biological Institute, University of Copenhagen, Universitetsparken 4, 2100 Kbh. Ø.

 

Fig. 1. Number (in boxes) of all charophyte all species (above) and Red-listed (below) in the four Scandinavian countries. Bray-Curtis similarity (in circles) derived from Red-list classifications of all species in pair-wise comparisons is shown at the national borders or along stippled lines connecting the regions. 

Eutrophication and acidification have reduced the number and quality of freshwater charophyte localities in Sweden, though not to the same extent as in Denmark apart from the SE-region, Scania, where agriculture is most intense (Blindow, 1992, 2000). Norway (NO, 324,220 km2) and Finland (FI, 337,030 km2) are located furthest to the north and lack charophyte species with northern distribution limits in Denmark and Sweden, while they have a few species with a predominantly northern or eastern distribution (in the case of Finland, Langangen, 2007). Both countries have experienced eutrophication and acidification, but like in Sweden high proportions of the land area are forests or

nature areas, though the calcareous regions most suitable for charophytes are cultivated as in the entire Scandinavia.

Denmark, Sweden and Finland have coastlines facing the brackish waters of the Baltic Sea which houses several charophyte species (Schubert & Blindow, 2003). These brackish populations have declined dramatically due to eutrophication of the southern Baltic Sea in Denmark, Germany and southern Sweden. The Swedish and Finnish coastlines in the central and northern parts of the Baltic Sea are less eutrophied and still support extensive, though diminishing, charophyte populations (Schubert & Blindow, 2003).

50

Niche specialization and traits of charophytes in Scandinavia  

  

Measuring abundance, diversity and

similarity

As a common measure of the relative species abundance of charophytes in all Scandinavian countries, we used the Red List categories (Langangen, 2007, Gärdenfors, 2010, Koistinen, 2010, Sjøtun et al., 2010). In the case of Denmark, we used a Red List proposal (Baastrup-Spohr, Dahl-Nielsen & Sand-Jensen, 2013a) based on a recent (1989-2010) compilation of the presence of charophytes in 226 water bodies.

Red List categories were translated to a numerical scale appropriate for statistical analysis: O = regionally extinct, 1 = critically threatened, 2 = endangered, 3 = vulnerable and 4 = nearly threatened. Species present in other Scandinavian countries but absent from a particular country were also given the value 0. Abundant species not present on the Red List were designated the value 6. We assume that these values approximate a logarithmic series of species abundances and confirmed this assumption by finding a highly significant (p < 0.001) linear relationship between the logarithm to number of localities (x) and the numerical Red List categories for both the largest (Sweden) and smallest country (Denmark) in the investigation (data extracted from Langangen, 2007, Gärdenfors, 2010, Baastrup-Spohr et al., 2013a) (Supplementary information, Fig. S1). Because of the high resemblance of numerical Red List categories vs. abundance relationship of the two countries (Fig. S1) and the identical criteria in the Red List regarding number of localities, we argue that the applied Red List values reliably reflect both national and Scandinavian abundance. Therefore, the abundance of individual species in the whole of Scandinavia was calculated as the sum of numerical Red List values in each of the four countries. Thus, the maximum possible abundance was 24 (= 4 x 6).

Diversity indices were calculated for the four Scandinavian countries by applying species number and species abundance as the

numerical values for species categories on and off the Red List as presented above. Richness was expressed as species number (S), evenness as the Pielou index (J´= H´/ln S) and the combination of richness and evenness as the Shannon-Wiener index (H´ = ∑ pi ln pi, where pi is the relative abundance of species according to numerical Red List categories). The similarity of abundances of species among the four Scandinavian countries was calculated by the Bray-Curtis (BC) index .

Niche parameters

There are multiple ways of measuring niche specialization each with their own advantages and drawbacks (Devictor et al., 2010). Here we used the Outlying Mean Index (OMI) analysis (Doledec, Chessel & Gimaret-Carpentier, 2000) providing a measure of niche position along multiple environmental gradients while searching for the most influential parameters. The advantage of OMI is that it makes no assumption concerning the shape of species response curves to the environment, and that it gives equal weight to species-rich and species-poor sites (Doledec et al., 2000).

The analysis results in a description of the mean niche position of each species, representing a measure of the distance between the mean conditions used by the species and the mean conditions of the study area. In our case, the resulting OMI values describe the distance of the niche for each charophyte species from the mean niche of a hypothetical charophyte occupying all investigated sites. The method has been successfully applied to terrestrial plants (Thuiller, Lavorel & Araujo, 2005) and in several freshwater environments (Doledec et al., 2000, Dole-Olivier, Malard, Martin et al., 2009). OMI analysis uses two matrices, a sampling unit vs. species matrix and sampling unit vs. niche variables.

We used environmental variables to delineate the species niche based on data collected by the Danish Environmental Agencies (Anonymous, 2012) from 1980 to

51

Niche specialization and traits of charophytes in Scandinavia  

  

2011 consisting of coupled registrations of species occurrence and environmental parameters (Table 1). In the case of closed freshwater and brackish water bodies with distinct in- and outlets, sites with both vegetation and environmental registrations were considered the sampling unit. In more open brackish and fully marine sites, the sampling units were marine environmental monitoring stations with nearby vegetation registrations. All vegetation registrations were collected within a 5 km distance of a monitoring station and in the same water body without any type of barrier (islands, isthmuses etc.). A few marine sites were sampled twice, with more than five years in between, if they showed signs of major changes in species composition over the period.

All registrations of species and associated environmental parameters were collected in the same year, except for a few instances where environmental data from the year immediately before or after species registrations were used. Environmental variables were calculated as time-weighted averages based on more than two measurements within the year.

We analyzed the niche position based on recordings in all investigated sites and in the freshwater and marine (> 0.5 g NaCl L-1) sites separately (Table 1).

Species traits

A series of species specific traits that could affect the species abundance was obtained from the literature on Scandinavian charophytes (Langangen, 2007) and supplemented by a Polish investigation in the few cases where Scandinavian data were missing (Gabka, 2009). The traits included were shoot height, life form, monoecious/dioecious and production of oospores and bulbils. The maximum shoot height (cm) was converted to a categorical scale: (0) 0-20 cm, (1) 21-40 cm and (3) >41 cm to account for the variation within and among species. Life form data were included as (0) obligate annual, (1) obligate perennial and (2) both annual and perennial or as (0) restricted life form (obligate annual or perennial) and (1) flexible life form (both annual and perennial). The same two data sources were used to determine whether species are monoecious (0) or dioecious (1). The intensity of oospore production was provided by Langangen (2007) for species with low (0) and high (2) production while species lacking detailed information were considered intermediary (1). Some species produce bulbils as a mean of vegetative reproduction and these species were given the value 1, while the rest were assigned the value 0.

Table 1. Range (min-max) of the variables included in the Outlying Mean Index (OMI) analysis performed at all sites, marine sites (>0.5 g NaCl L-1) and freshwater sites. The variables were collected very close to (freshwater sites) or within 5 km of vegetation registration (marine sites). Data source is the Danish Environmental Monitoring Program. N/A is variables not included in the calculations of OMI for a given dataset.

Variables Full dataset Marine sites Freshwater sites Salinity (g NaCl L-1) 0-26.7* 0.6-26.7 N/A Chlorophyll a (µg L-1) 0.9-295.2 0.9-295.2 1.1-127.8 Total phosphorus (µg L-1) 7.6-863.6 18.8-407.5 7.6-863.6 Total nitrogen (µg L-1) 244.9-5729.3 244.9-5729.3 286.4-4982.1 Alkalinity (meq L-1) 0.025-5.6 Secchi depth (m) 0.26-7.8 Lake depth maximum (m) 0.2-32.7 Lake area (ha) 0.17-1730**

*In all freshwater sites salinity was very low and measured infrequently and all freshwater sites were given the value 0 in the full dataset and the variable was not included in the freshwater OMI analysis. **Lake area was log transformed in the OMI analysis.

52

Niche specialization and traits of charophytes in Scandinavia  

  

Species salinity tolerance was collected from Schubert and Blindow (2003), and was included as the range in salinity (max-min) inhabited by a given species. The brackish water species, Tolypella normaniana, was not included in the analysis as tolerance values could not be obtained from the literature. A summary of the values of the investigated trait and the Scandinavian abundance is provided in the supplementary material (Table S2) Results

Species richness and similarity

Thirty five species have been found in Scandinavia within the last 50 years. In historical time, a few more species such as Chara baueri and Nitella tenuissima have been recorded but are now considered regionally extinct. Species richness of charophytes was highest in Sweden (32) and lower in the other countries (20-25, Fig. 1). A large number of species were on the Red List in all countries (11-17, Fig. 1) corresponding to 50-87% of all species. The Shannon Wiener biodiversity index, combining richness and evenness, was highest in Sweden (3.38) and lower in Norway (3.07), Denmark (3.00) and Finland (2.93) mainly due to changes in species richness. The Pilou evenness index was about the same in all four Scandinavian countries (0.958-0.977).

Similarity in relative abundance of charophytes between the countries was a close reflection of their proximity (Fig. 1). Similarity was high between neighboring countries and declined with distance between countries.

Niche and abundance

The OMI analysis used registrations of charophytes in 226 Danish sites with a total of 19 species. Omitting one species only registered once (Chara rudis) and using the restricted number of variables sampled across all marine and freshwater sites yielded an OMI first axis explaining 67% of the observed variability with high scores for

salinity in particular (Table 2). As several environmental variables were only measured in the freshwater sites, separating the dataset into a marine and a freshwater section allowed a more detailed description of the relative niche position of the freshwater species. Both in the freshwater and marine sites the OMI first axis explained a high percentage of the variability (Table 2). In the marine dataset, the niches separated primarily along a main gradient of salinity, negatively coupled to nutrients and chlorophyll a (Table 2). In the analysis of the freshwater sites the separation of species niches was linked to two gradients; one in alkalinity and a second in lake area. This indicates that niches of freshwater species separate along gradients (e.g. alkalinity) that are not relevant in the permanently alkaline marine waters.

To test if the abundance of charophytes in Scandinavia was related to niche specialization, we correlated the abundance with the OMI index values of each species. Using the niche specialization from the full dataset we found a tendency towards a negative relationship between niche specialization and Scandinavian abundance (r=-0.41, p=0.09) (Fig. 1). Table 2. Results of the Outlying Mean Index analyses (OMI) performed on the full dataset, marine (>0.5 g NaCl L-1) and freshwater sites. Scores of the included variables and the amount of variability explained are given for the first two axes. Full dataset Marine

sites Freshwater

sites Axes 1 2 1 2 1 2 % of variability 78 17 86 10 40 33 Salinity (g NaCl L-1)

0.78 -0.11 0.61 -0.31

Chlorophyll a (µg L-1)

-0.10 -0.25 -0.54 -0.17 -0.18 -0.05

Total phosphorus (µg L-1)

-0.10 -0.17 -0.51 -0.13 -0.04 -0.17

Total nitrogen (µg L-1)

-0.19 -0.22 -0.56 -0.06 -0.10 -0.16

Alkalinity (meq L-1)

-0.37 -0.07

Secchi depth (m) 0.07 0.07 Lake depth maximum (m)

-0.03 0.06

Lake area (ha) -0.17 0.36

53

Niche specialization and traits of charophytes in Scandinavia  

  

However, separating freshwater and marine sites, allowing inclusion of variables important to niche separation in freshwaters, yielded a significant (p< 0.05) relationship between Scandinavian abundance and niche specialization (Fig. 2). We thus found robust evidence supporting the hypothesis that abundant charophytes are less specialized than rarer species. Traits and abundance

The abundance of Scandinavian charophytes was significantly related to several of the investigated species traits. Taller species tended to be more abundant than small species which was also the case for species with wide salinity tolerance (Table 3).

Annual species were significantly less abundant than species capable of expressing both an annual and perennial life form while obligatory perennials grouped in between (Kruskall-Wallis, p= 0.03). However, merging the two groups of restricted life forms (obligate annual and perennial) and comparing them to species with flexible life form shoved that species with flexible life form were generally more abundant (Mann-Whitney, p= 0.03). The intensity of spore production and presence of bulbils did not affect the abundance. However, species bearing bulbils tended to have wider salinity tolerance, lower sexual reproductive investment and to be dioecious (Table 3).

Discussion

Diversity and similarity

Our results confirmed the prediction that the proportion of species on the Red List is high in individual Scandinavian countries (50-87%) and this prediction can be extended to Germany and countries on the Balkan Peninsula (Schmidt et al., 1996, Blazencic et al., 2006) emphasizing the highly critical status of charophytes here and elsewhere in Europe (Simons & Nat, 1996, Auderset Joye et al., 2002).

These high proportions of species on the Red List are unprecedented among aquatic submerged plants in Denmark and Norway (30-35%) and terrestrial plants in all four Scandinavian countries (18-28%) (Gärdenfors, 2010, Kålås et al., 2010, Koistinen, 2010, Wind & Pihl, 2010).

Fig. 2. The relationship between niche specialization and Scandinavian abundance. The niche specialization was calculated as the Outlying Mean Index (OMI) using specific environmental conditions for Danish growth sites. OMI values are shown as % of inertia. The niche specialization was calculated based on three datasets, using all growth sites (full dataset), marine (>0.5 g NaCl L-1) growth sites and only freshwater sites. The lines shown are model II regressions. The relationships between abundance and OMI are significant for the freshwater and marine datasets (Spearman, p< 0.05).

0

20

40

60

80

100

Fre

shw

ater

OM

I (%

)

0

20

40

60

80

100M

arin

e O

MI

(%)

0 10 20 300

20

40

60

80

100

Scandinavian abundance

Ful

l dat

aset

OM

I (%

)

54

Niche specialization and traits of charophytes in Scandinavia  

  

We explain the highly critical status of charophytes by the fact that relatively few and small habitats suitable for charophytes are available nowadays in Scandinavia because of extensive habitat loss and deterioration by eutrophication, acidification, and mechanical perturbation (Baastrup-Spohr et al., 2013b). Moreover, charophytes are much more sensitive to eutrophication than aquatic plants because they are competitively inferior and occupy habitats below the canopy of tall plants and at the greatest depth of submerged vegetation being highly sensitive to loss of vegetation (Blindow, 1992, Van den Berg et al., 1998). For example, when Lake Fure became eutrophied and turbid from 1940 to 1985, all 11 charophyte species were lost whereas only 8 of the original 21 rooted plant species vanished (Sand-Jensen et al., 2008).

The number of charophyte species observed within a country is a poor descriptor of their status because many species are only observed in very few localities (Auderset Joye et al., 2002, Blazencic et al., 2006). Although the number of charophyte localities has declined very markedly in Scandinavia during the last 100 years most species previously recorded by Olsen (1944) have persisted (Baastrup-Spohr et al., 2013b). A few species have even been added in some countries in recent years due to either establishment of new species, acceptance of new species following taxonomic reevaluation or because of more intensive field work (Langangen, 2007). However, the great loss of suitable

freshwater and brackish sites due to heavier eutrophication, drainage and mechanical deterioration in Denmark account for the loss of species abundance (Baastrup-Spohr et al., 2013b).

Abundance, specialization and traits

As predicted, charophytes with low niche specialization were more abundant in Scandinavia while specialist species were rarer. This relationship will exist if species that have broad environmental tolerances and are able to use a wide variety of resources will in so doing achieve high local densities and be able to survive in more places over larger areas (Gaston & Spicer, 2001).

Some investigators argue that it is a main analytical problem in studies of abundance versus range size that species are evaluated based on their realized niche, while the fundamental niche is best determined experimentally. Experimental determinations of the fundamental niche breadth with respect to several biological features are few because of tradition and insufficient experimental expertise among investigators. However, even thorough experimental studies, if available, may not provide the necessary answers because they are restricted in time and space, include a limited number of traits and not necessarily the essential ones and never include the greater variety of genotypes that are supposed to be present in widely distributed common species.

Table 3. Relationships between charophyte abundance in Scandinavia and investigated species traits. Also shown, are the within species correlation of the functional traits. The given values are Spearman rank correlations with the p-value in parenthesis.

Trait Scandinavian abundance

Height Salinity tolerance

Oospore production

Bulbil production

Life form

Height 0.46 (<0.01) Salinity tolerance 0.59(<0.01) 0.16 (0.34) Oospore production 0.05 (0.74) -0.27 (0.11) 0.11 (0.54) Bulbil production 0.35 (0.03) 0.36 (0.03) 0.48

(<0.01) -0.39 (0.02)

Life form 0.40 (0.02) 0.40 (0.02) 0.28 (0.12) -0.11 (0.54) 0.16 (0.37) Monoecious/diocious 0.12 (0.48) 0.05 (0.77) 0.29 (0.09) -0.20 (0.24) 0.36 (0.03) -0.04 (0.82)

55

Niche specialization and traits of charophytes in Scandinavia  

  

Among charophytes, for example, the common species Chara vulgaris is highly polymorphic with at least five noticeable varieties in Scandinavia which may ensure greater ecophysiological variability and permitting the species to thrive in many places (Langangen, 2007). It is highly unlikely that all five varieties are ever accounted for in the description of the fundamental niche. Therefore, there is no guarantee that experiments disclose the relevant fundamental niche.

The historical landscape development in, particularly, southern Scandinavia, has been characterized by widespread eutrophication. Such a development can lead to biotic homogenization with a decrease of specialists and an increase of generalists (McKinney & Lockwood, 1999). For charophytes such as Chara globularis, C. vulgaris and Nitella flexillis that have been shown to increase in, or retain high abundance within the last century in Denmark, The Netherlands and Switzerland all have low values of specialization (Simons & Nat, 1996, Auderset Joye et al., 2002, Baastrup-Spohr et al., 2013b). In contrast, highly specialized species such as Laprothamnium papulosum has declined dramatically (Schubert & Blindow, 2003).

Recent studies on terrestrial plant species suggest that their success is intimately linked to the functional traits (Mokany & Roxburgh, 2010). Charophytes of the Scandinavian countries support this pattern because species abundance was significantly linked to several of the investigated traits. Shoot height of charophytes as a proxy for competitive capability for light and space in dense stands was positively linked to Scandinavian abundance, similar to findings for vascular Potamogeton species (Sand-Jensen et al., 2000). This result indicates that the ability to compete for light plays an important role in shaping the charophyte community. The ability to physiologically tolerate a large range of salinities should also be of great advantage in Scandinavia because of the

proximity and close linkage between marine and freshwater habitats and the mere increase of the number of potential growth sites. This was verified by the relationship between abundance and salinity tolerance, implying that species with wide tolerances were generally more abundant than species with smaller ranges. For some species (e.g. Chara aspera), distribution maps reveal that luxuriant growth in certain coastal waters is accompanied by the presence of species in inland waters nearby (Olsen, 1944, his Fig. 31). Interestingly, the salinity tolerance seems to co-vary within species with the formation of bulbils suggesting that this mode of reproduction is more essential for dispersal and survival of species in brackish waters than freshwater bodies. Apomictic reproduction has been shown to be more abundant in harsh environments (Richards, 2003). Brackish environments with changing salinity and strong physical disturbance (Idestam-Almquist, 2000) could be regarded as harsher than freshwaters giving rise to more apomictic species.

Abundance of charophyte species in Scandinavia was also positively correlated to the existence of both annual and perennial rather than only one life form within species. This effect could be straightforward because the breath of life forms increases the ability of species to thrive both in perturbed habitats as annual forms and in stabile habitats as perennial forms. Charophytes develop dense monospecific stands as winter annuals in ephemeral ponds that dry out during summer thanks to efficient survival of numerous oospores and sprouting during refilling in autumn. Vascular aquatic plants are impeded much more than charophytes by summer desiccation and their contribution increases gradually from temporary to permanent calcareous ponds (Sand-Jensen et al., 2010). In permanent ponds and lakes, summer annual forms may thrive during summer in shallow waters and survive heavy physical disturbance and thick ice cover during the winter as oospores and vegetative propagules. Perennial

56

Niche specialization and traits of charophytes in Scandinavia  

  

forms may grow year-round in the physically more stable deeper waters where more time is needed for population growth because of low light availability. Certain charophytes have an exceptional capability of growing at lower light and temperature than vascular plants in the cold bottom waters of temperate lakes (Nygaard & Sand-Jensen, 1981).

A diagnosis of being common or rare

Our data confirmed that charophyte’s traits and degree of niche specialization are important determinants of whether species are common or rare. In common species the prevalent traits are related to being large and having a flexible life form and a high salinity tolerance. In rare species life traits are associated with having more specialized niches.

Some rare charophyte species have derived from quite common species which they still resemble. Thus, Chara filiformis resembles the common C. contraria and Nitella capillaris resembles the common N. opaca (Langangen, 2007). With reference to the rich terrestrial Cape flora with many endemics, Davies et al. (2011) proposed that if plants speciate via small isolated populations at the edge of larger species ranges, then lineages that are diversifying rapidly will have many threatened species. Already Darwin (1859) noted the pattern of widespread, variable plant species acting as species pumps via peripheral sister species of restricted distribution. If this mechanism also operates in charophytes, then formation of peripheral species in cold regions of Scandinavia could be driven by widespread, common species. If we fragment the ranges of these common species they will lose their ability to operate as species pumps in the future (Knapp, 2011).

In conclusion, most charophytes are rare in Scandinavia because of too few suitable habitats and high niche specificity. In contrast to early believes, charophytes are not solely restricted to stabile, resource-poor habitats (ultra-oligotrophy or deep waters of low irradiance). Many charophytes flourish in

recently constructed ponds and restored lakes at relatively high nutrient levels before phytoplankton or tall vascular plants eventually take over and outcompete them (Crawford, 1977, Wade, 1990, Baastrup-Spohr et al., 2013b). The best way to set long-term priorities for conservation of charophyte diversity is to offer a flow of new pond habitats, ensure diverse habitats with clean water and avoid physical deterioration for the benefit of both common and rare species.

Acknowledgements

We thank the Villum Kann Rasmussen Centre of Excellence, Centre for Lake Restoration (CLEAR) for financial support, Dr. Dorte Krause-Jensen for providing marine botanical data and Drs. Irmgaard Blindow and Hans Henrik Kehlet Bruun for valuable comments to the manuscript. We are indebted to many Danish field botanists for information on the distribution of charophytes. References

Anonymous (2012) Danish environemtal portal. http://www.miljoeportal.dk/.

Auderset Joye, D., Castella, E. & Lachavanne, J.B. (2002) Occurrence of Characeae in Switzerland over the last two centuries (1800-2000). Aquatic Botany, 72, 369-385.

Baastrup-Spohr, L., Dahl-Nielsen, J. & Sand-Jensen, K. (2013a) Rare and threatened charophytes (in Danish). Urt, in review.

Baastrup-Spohr, L., Iversen, L.L., Dahl-Nielsen, J. & Sand-Jensen, K. (2013b) Seventy years of changes in the abundance of Danish charophytes. Freshwater Biology, Accepted.

Biggs, J., Williams, P., Whitfield, M., Nicolet, P. & Weatherby, A. (2005) 15 years of pond assessment in Britain: results and lessons learned from the work of Pond Conservation. Aquatic Conservation-Marine and Freshwater Ecosystems, 15, 693-714.

57

Niche specialization and traits of charophytes in Scandinavia  

  

Blazencic, J., Stevanovic, B., Blazencic, Z. & Stevanovic, V. (2006) Red data list of charophytes in the Balkans. Biodiversity and Conservation, 15, 3445-3457.

Blindow, I. (1992) Decline of charophytes during eutrophication - comparison with angiosperms. Freshwater Biology, 28, 9-14.

Blindow, I. (2000) Distribution of Charophytes along the Swedish coast in relation to salinity and eutrophication. International Review of Hydrobiology, 85, 707-717.

Blindow, I., Dietrich, J., Mollmann, N. & Schubert, H. (2003) Growth, photosynthesis and fertility of Chara aspera under different light and salinity conditions. Aquatic Botany, 76, 213-234.

Bonis, A. & Grillas, P. (2002) Deposition, germination and spatio-temporal patterns of charophyte propagule banks: a review. Aquatic Botany, 72, 235-248.

Brochet, A.L., Guillemain, M., Gauthier-Clerc, M., Fritz, H. & Green, A.J. (2010) Endozoochory of Mediterranean aquatic plant seeds by teal after a period of desiccation: Determinants of seed survival and influence of retention time on germinability and viability. Aquatic Botany, 93, 99-106.

Brown, J.H. (1984) On the relationship between abundance and distribution of species. American Naturalist, 124, 255-279.

Clavel, J., Julliard, R. & Devictor, V. (2011) Worldwide decline of specialist species: toward a global functional homogenization? Frontiers in Ecology and the Environment, 9, 222-228.

Crawford, S.A. (1977) Chemical, physical and biological changes associated with Chara succession in farm ponds. Hydrobiologia, 55, 209-217.

Darwin, C. (1859) On the origin of species by means of natural selection, John Murray and Sons, London.

Davies, T.J., Smith, G.F., Bellstedt, D.U., Boatwright, J.S., Bytebier, B., Cowling,

R.M., Forest, F., Harmon, L.J., Muasya, A.M., Schrire, B.D., Steenkamp, Y., Van Der Bank, M. & Savolainen, V. (2011) Extinction risk and diversification are linked in a plant biodiversity hotspot. Plos Biology, 9.

Devictor, V., Clavel, J., Julliard, R., Lavergne, S., Mouillot, D., Thuiller, W., Venail, P., Villeger, S. & Mouquet, N. (2010) Defining and measuring ecological specialization. Journal of Applied Ecology, 47, 15-25.

Dole-Olivier, M.J., Malard, F., Martin, D., Lefebure, T. & Gibert, J. (2009) Relationships between environmental variables and groundwater biodiversity at the regional scale. Freshwater Biology, 54, 797-813.

Doledec, S., Chessel, D. & Gimaret-Carpentier, C. (2000) Niche separation in community analysis: A new method. Ecology, 81, 2914-2927.

Futuyma, D.J. & Moreno, G. (1988) The evolution of ecological specialization. Annual Review of Ecology and Systematics, 19, 207-233.

Gärdenfors, U. (2010) The 2010 red list of Swedish species. Artsdatabanken, SLU, Upsala.

Gaston, K.J. & Spicer, J.I. (2001) The relationship between range size and niche breadth: a test using five species of Gammarus (Amphipoda). Global Ecology and Biogeography, 10, 179-188.

Idestam-Almquist, J. (2000) Dynamics of submersed aquatic vegetation on shallow soft bottoms in the Baltic Sea. Journal of Vegetation Science, 11, 425-432.

Kålås, J.A., Viken, Å., Henriksen, S. & Skjelseth, S. (2010) The 2010 Norwegian Red List for Species, Norwegian Biodiversity Information Centre.

Keddy, P.A. (1992) Assembly and response rules - two goals for predictive community ecology. Journal of Vegetation Science, 3, 157-164.

58

Niche specialization and traits of charophytes in Scandinavia  

  

Knapp, S. (2011) Rarity, species richness, and the threat of extinction - are plants the same as animals? Plos Biology, 9.

Koistinen, M. (2010) Stoneworts, Characeae. In: The 2010 Red List of Finish Species. (Ed^Eds P. Rassi & E. Hyvärinen & A. Juslén & I. Mannerkoski). Ympäristöministeriö & Suomen ympäristökeskus, Helsinki.

Krause, W. (1997) Charales (Charophyceae). In: Süsswasserflora von Mitteleuropa. (Ed^Eds H. Ettl & G. Gärtner & H. Heynig & D. Mollenhauer). Gustav Fisher, Stuttgart.

Langangen, A. (2007) Charophytes of the Nordic countries, Saeculum ANS, Oslo.

Mckinney, M.L. & Lockwood, J.L. (1999) Biotic homogenization: a few winners replacing many losers in the next mass extinction. Trends in Ecology & Evolution, 14, 450-453.

Mokany, K. & Roxburgh, S.H. (2010) The importance of spatial scale for trait-abundance relations. Oikos, 119, 1504-1514.

Nygaard, G. & Sand-Jensen, K. (1981) Light climate and metabolism of Nitella flexilis (L) Ag in the bottom waters of oligotrophic Lake Grane Langsø, Denmark. Internationale Revue der gesamten Hydrobiologie, 66, 685-699.

Olsen, S. (1944) Danish Charophyta, Det Kongelige Danske Videnskabernes Selskab, Copenhagen.

Richards, A.J. (2003) Apomixis in flowering plants: an overview. Philosophical Transactions of the Royal Society of London Series B-Biological Sciences, 358, 1085-1093.

Riis, T. & Sand-Jensen, K. (1998) Development of vegetation and environmental conditions in an oligotrophic Danish lake over 40 years. Freshwater Biology, 40, 123-134.

Sand-Jensen, K., Baastrup-Spohr, L., Møller, C.L., Winkel, A., Brodersen, K.P., Borum, J. & Staehr, P.A. (2010) Plant distribution patterns and adaptation in a limestone

quarry on Öland. Svensk Botanisk Tidsskrift, 104, 23-31.

Sand-Jensen, K., Pedersen, N.L., Thorsgaard, I., Moeslund, B., Borum, J. & Brodersen, K.P. (2008) 100 years of vegetation decline and recovery in Lake Fure, Denmark. Journal of Ecology, 96, 260-271.

Sand-Jensen, K., Riis, T., Vestergaard, O. & Larsen, S.E. (2000) Macrophyte decline in Danish lakes and streams over the past 100 years. Journal of Ecology, 88, 1030-1040.

Schmidt, D., K., W.V.D., Krause, W., Kies, L., Garniel, A., Geissler, U., Gutovski, A., Samietz, R., Schütz, W., Vahle, H.-C., Vöge, M., Wolff, P. & Melzer, A. (1996) Rote Liste der Armleuchteralgen (Charophyceae) Deutschlands, Schriftenreihe für Vegetationskunde.

Schubert, H. & Blindow, I. (2003) Charophytes of the Baltic Sea, Koeltz Scientific Books, Koenigstein.

Shipley, B. (2010) Community assembly, natural selection and maximum entropy models. Oikos, 119, 604-609.

Simons, J. & Nat, E. (1996) Past and present distribution of stoneworts (Characeae) in the Netherlands. Hydrobiologia, 340, 127-135.

Sjøtun, K., Fredriksen, S., Heggøy, E., Husa, V., Langangen, A., Lindstrøm, E., Moy, F., Rueness, J. & Åsen, P.A. (2010) Cyanophyta, Rhodophyta, Chlorophyta, Ochrophyta. In: The 2010 Norwegian Red List for Species. (Ed^Eds J.A. Kålås & Å. Viken & S. Henriksen & S. Skjelseth). Norwegian Biodiversity Information Centre, Trondheim.

Thuiller, W., Lavorel, S. & Araujo, M.B. (2005) Niche properties and geographical extent as predictors of species sensitivity to climate change. Global Ecology and Biogeography, 14, 347-357.

Torn, K., Martin, G., Kotta, J. & Kupp, M. (2010) Effects of different types of mechanical disturbances on a charophyte dominated macrophyte community.

59

Niche specialization and traits of charophytes in Scandinavia  

  

Estuarine Coastal and Shelf Science, 87, 27-32.

Van Den Berg, M.S., Coops, H., Simons, J. & De Keizer, A. (1998) Competition between Chara aspera and Potamogeton pectinatus as a function of temperature and light. Aquatic Botany, 60, 241-250.

Wade, P.M. (1990) The colonisation of disturbed freshwater habitats by Characea. Folia Geobotanica and Phytotaxonomica, 25, 275-278.

Wind, P. & Pihl, S. (2010) The Danish Red List. The National Environmental Research Institute, Aarhus University http://redlist.dmu.dk.

Winter, U., Soulie-Märsche, I. & Kirst, G.O. (1996) Effects of salinity on turgor

pressure and fertility in Tolypella (Characeae). Plant Cell and Environment, 19, 869-879.

Supporting information

The supporting information is found in final part of the PhD-thesis. Fig. S1. Comparison between relationship between Red List status and observed abundance in Sweden and Denmark. Both relationships are highly significant and the two slopes not significantly different. Table S2. The abundance and functional traits of each charophytes species found in Scandinavian waters.

60

61

 

62

Community assembly from pond to alvar

From pond to alvar: predicting plant community composition along a steep hydrological gradient Lars Baastrup-Spohr1*, Kaj Sand-Jensen1, Sascha Veggerby Nicolajsen2 and Hans Henrik Bruun2

1Freshwater Biological Laboratory, Dept Biology, University of Copenhagen, Helsingørsgade 51, 3400 Hillerød, Denmark; 2Center for Macroecology Evolution and Climate, Dept Biology, University of Copenhagen, Universitetsparken 15, 2100 København Ø ________________________________________________________________________________ Summary 1. Why do plants grow where they grow? Prediction of species’ occurrence and abundance in relation to the environment is a core aim to ecology and so is understanding the link between environmental stressors and adaptive traits. Conceptually, community assembly may be viewed as a sequence of filters, sieving off species according to their functional traits and thus restricting the range of trait values within a community. The filtering concept is the theoretical basis of a maximum entropy (maxent) model developed to predict species’ abundances in communities. 2. We investigated plant community composition and six morpho-physiogical plant traits along a pronounced hydrological gradient in an abandoned limestone quarry. We studied the strength of environmental filtering on traits and the importance of the different traits to community assembly, investigating whether the traits experiencing filtering also prove valuable in predicting community composition. 3. We show that root porosity is significantly filtered by the environment along a marked hydrological gradient, as is specific leaf area and resistance to water loss on drying. For individual traits, the strength of filtering wax and wane along the gradient, strongly suggesting that the mechanisms through which species are filtered into communities act through different traits as environmental conditions change along the gradient. 4. The maxent model predicted 66% of the variation in species’ relative abundances using six traits. The traits subject to filtering also were most important in predicting species abundance. 5. Synthesis. Our results show that few plant traits are exposed to environmental filtering across the entire gradient in water coverage and soil depth, and that most traits are subject to strong filtering only in parts of the gradient. Evidence for congruence between trait dispersion indices and the maxent model was established, underpinning the importance to plant community assembly of environmental filtering of species through their traits. Key-words Alvar, community assembly, environmental filtering, maximum entropy, Öland, plant trait, species richness, trait-environment relationship.

63

Community assembly from pond to alvar

Introduction An important goal of community ecology is to predict the occurrence and abundance of species inhabiting natural communities (McGill et al., 2006). Attaining this goal requires thorough appreciation of the mechanisms important to community assembly. These mechanisms have been likened to a series of filters determining which species from the regional species pool can complete their life cycle under the prevailing abiotic and biotic conditions (Keddy, 1992, Grime, 2006). The view that filters are operating on functional traits of species - rather than on species themselves, has received great attention recently (McGill et al., 2006, Shipley et al., 2006, Ackerly & Cornwell, 2007, Cornwell & Ackerly, 2009), although the idea that the distribution of species must be explained through their functional traits is deeply rooted in ecology (e.g. Warming, 1895).

While the abiotic and biotic environment selects for or against certain plant traits, the traits themselves may determine species interactions and ecosystem function (Grime, 1998, Hooper et al., 2005, McGill et al., 2006, Lavorel, 2013). Along environmental gradients, variability of the traits expressed in communities change and these changes are associated with shifting rules of community assembly (Cornwell & Ackerly, 2009, Bernard-Verdier et al., 2012). A smaller range in traits values than expected from randomness – trait clustering – may be taken as an indication of trait filtering, because species having trait values outside a given allowed trait range succumb (Keddy, 1992, Cornwell & Ackerly, 2009). Sorting of species into communities appears to be particularly distinct along environmental gradients between land and freshwaters, which originally sparked the idea of environmental filters (van der Valk, 1981). Here, in particular water coverage (flooding) and sediment conditions (e.g. organic content and anoxia) change over short distances and strongly affect the abundance of plant species

and composition of communities (van der Valk, 1981, Silvertown et al., 1999, van Eck et al., 2004). This background stimulated us to study the importance of abiotic factors for the filtering of traits and how this influences the ability to predict species abundance in an area of particularly steep gradients in water coverage and sediment depth along a gradient from permanent ponds to dry shallow soils on the limestone alvar of Öland, SE Sweden.

Water coverage affects plants through the 104-fold reduction of gas diffusion rates and the 20-25-fold lower solubility of oxygen at saturation compared to atmospheric air (Crawford, 2009). These basic physical differences impose severe constrains on the supply rates of inorganic carbon for photosynthesis and oxygen for respiration of plants in water relative to air (Madsen & Sand-Jensen, 1991). Plants in submerged or water-logged soils, often have gas filled tissue, to alleviate problems of poor oxygen supply (Justin & Armstrong, 1987, Končalová, 1990). Aerenchyma in roots is essential to transport of oxygen from shoots to the tips of root buried in anoxic sediments (Colmer, 2003). Aerenchyma formation increases substantially both within and between species distributed from dry and well oxygenated soils to water-logged, anoxic soils (Mommer et al., 2006). While submerged plants never experience water shortage, the water available to terrestrial plants tends to decrease with decreasing soil depth (Pérez-Ramos et al., 2012). Along with the decrease in soil moisture, nutrient availability will also decline.

Along a gradient in soil depth and water availability traits such as the wilting point of shoots and the rate of water loss at high temperature, measures of the ability to endure conditions with insufficient external water supply, are expected to respond. The ability to resist water loss on drying (WLOD) should increase from a minimum in deep soils of high water holding capacity to a maximum in shallow soils and it should decline with water coverage as submerged species lack

64

Community assembly from pond to alvar

protection against desiccation (Wetzel, 2001) while in occasionally flooded areas both strategies could co-oocur leading to increased trait range. Also specific leaf area (SLA) has been shown to decrease along gradients in water availability (Violle et al., 2011, Bernard-Verdier et al., 2012). Among permanently submerged aquatic plants (Nielsen & Sand-Jensen, 1989) and temporarily submerged amphibious or terrestrial plants (Mommer & Visser, 2005), the restricted inorganic carbon supply to photosynthetic tissues has resulted in different strategies, such as high-SLA, often serrate, leaves to maximise the surface area to volume ratio. Other leaf traits such as dry matter content (LDMC) and surface area are also related to soil moisture in terrestrial environments (Hallik et al., 2009, Violle et al., 2011, Bernard-Verdier et al., 2012) and they are likely to change with the frequency of flooding. The different strategies of aquatic plants concerning formation of submerged, floating and aerial leaves may, however, distort this picture by causing trait over-dispersion (Bernard-Verdier et al., 2012).

A methodology for prediction of the relative abundance of species in communities was formulated by Shipley and co-workers (2006) using a maximum entropy model constrained by community aggregated trait values. The model´s concept is that if species filtering processes constrain the composition of a local community in such a way that the individual species possessing suitable functional traits become more abundant than predictable from all regional species available then “community-level properties should emerge” (Laliberté et al., 2012). An important assumption when modelling with community aggregated traits as constraints is that these traits are actually subject to environmental filtering. Trait filtering by the environmental conditions should restrict the range of values of a given trait at a given site and is often evaluated by comparison to what could be expected due to random processes (Cornwell

& Ackerly, 2009, Ingram & Shurin, 2009, Bernard-Verdier et al., 2012).

We aim at evaluating whether traits, which are subject to strong environmental filtering, also proves to work in predicting the presence and abundance of species in communities distributed along a hydrological gradient. We do this by testing the following three specific hypotheses: 1) In local communities, using few but biologically relevant functional traits, we can make strong prediction of which plant species from a species pool will be found in which relative abundances, 2) The strength of environmental filtering of traits will change along a steep environmental gradient according to the selective pressure imposed, and 3) Functional traits experiencing environmental filtering are strong predictors in models of community assembly. For the tests, we used the framework of Ingram & Shurin (2009) and Shipley et al. (2006) on a coupled community-trait-environment data set collected in an abandoned limestone quarry surrounded by intact Alvar vegetation on the Baltic island of Öland, Sweden.

Material & Methods Site The investigation was conducted in an abandoned limestone quarry on Räpplinge Alvar on the island of Öland, SE Sweden (Sand-Jensen & Jespersen, 2012). The substratum consists of exposed solid limestone pavements, which are either extremely dry and almost devoid of vegetation over large areas, covered by shallow soil or sediment layers (up to a few cm) or covered by more permanent vegetation in slightly deeper soils and sediments in wetter places, in cracks between the limestone plates or in depressions close to and within small ponds. The quarry was abandoned about 30 years ago and has been managed by horse grazing (2 horses on 6 ha in summer) keeping the entire area open and mostly without shrubs and trees.

65

Community assembly from pond to alvar

Table 1. Environmental conditions measured in analyses of trait-environment relationships in an abandoned limestone quarry on Öland, Sweden. Environmental variates 1-5 are calculated from continuous water level registration and assesses the conditions of flooding and drought. Variates 6-14 are based on measurements within each community sample plot.

Environmental parameter Number Description Range Days with water cover 1 Number of days with a positive water value 0-51 days Longest water cover 2 Longest period of positive water values 0-51 days Days with drought 3 Number of days with a negative water value <-

5cm 0-51 days

Summed depth 4 Sum of all positive water values 0-1969.1 cm Summed water value 5 Sum of all water values -694.7-1969.1 cm Water depth in plot 6 Manually measured water depth in each plot

0-35 cm

Sediment depth 7 Mean sediment/soil depth 0.7-16 cm Standard deviation of sediment depth

8 Standard deviation of soil depth 0.5-7.6 cm

Vegetation height 9 Mean vegetation height 1.7 -36.6 cm Standard deviation of vegetation height

10 Standard deviation of vegetation height 0.6-22.9 cm

Abundance of moss 11 Presence of moss in subplots 0-100% Abundance of pebble 12 Presence of pebble in subplots 0-100% Abundance of bare soil/sediment

13 Presence of bare soil/sediment in subplots 0-100%

Abundance of Nostoc commune

14 Presence of N. commune in subplots 0-100%

The local climate is quite dry (mean annual precipitation 510 mm; 1960-1990), with cold winters (January mean -1.2o C) and relatively mild summers (July mean 16.2o C) (SMHI, 2013). The precipitation is evenly distributed throughout the year (monthly mean range 32-54 mm), but temperature variations lead to large differences in the evapotranspiration and thus water availability (SMHI, 2013). Between April and August in 2010, surface temperature on the exposed limestone pavements exceeded 40o C on 37 days (21%) and reached a maximum of 50o C on two days in July (Sand-Jensen & Jespersen, 2012). Thus, the area is characterized by extreme gradients in water availability between the permanent water-filled ponds and the exposed limestone pavements with mm-thin layers of dust where water is only available during and immediately after wet weather (Sand-Jensen et al., 2010). The drainage water from the limestone soils filling the shallow ponds has a substantial acid neutralizing capacity (2 meq l-1) and a pH of 8.2 at air saturation and pH 7 at 30-fold super-saturation typical of soils (Sand-Jensen & Staehr, 2012, Christensen et

al., 2013). The soils in the investigated site had low phosphorus concentrations (0.1-0.14 µg TP g DW-1) and the ponds had very low concentrations of soluble inorganic nitrogen and phosphorus (Christensen et al., 2013). Some representative species are Sedum album L. occurring abundantly in the driest places, Juncus articulatus L. and Alopecurus geniculatus L. growing abundantly in the bank zone of ponds, and Chara aspera C.L.Willdenow being common in the ponds.

Vegetation sampling and water regime The vascular vegetation was analysed quantitatively between 23 and 28th of May 2010 in twenty-one 0.25 m2 squares (referred to as local communities) placed by stratified random sampling to encompass the different soil depths, water regimes and vegetation physiognomies within a 100 by 200 m area. Every local community was located close to one of four focal ponds in the area. All plants except seedlings were identified to species and recorded by their presence or absence in each of twenty-five 10 10 cm subplots within the 0.25 m2 square. A total of 525

66

Community assembly from pond to alvar

subplots were examined and 65 species were recorded.

Ground surface without vascular plants was classified as covered by moss, colonial cyanobacterium Nostoc commune Vaucher, pebble or bare soil and the presence of each type was recorded in the subplots. Maximum vegetation height (cm), water depth (cm) and soil depth (cm) were measured in each subplot (soil depth by poking a thin metal peg vertically into the soil (Table 1).

The highly dynamic water regime was presumed to be of utmost importance for plant community assembly due to infrequent precipitation events, shallow soils, impermeability of the limestone, elevation heterogeneity and the hot microclimate of the study site. Therefore, we calculated the daily water depth in each plot by placing two continuously operating barometric pressure sensors along with an atmospheric reference (HOBO U20, Onset, Massachusetts, USA) in two of four ponds in the area in the beginning of the growing season 2010 (April 8). In the two ponds without continuous recordings of water level, four manual recordings were made during the season, and the water level dynamics of these two ponds corresponded very closely to one of the continuously measured ponds. We thus modelled the water level in these two ponds by normalising the continuous measurements of this pond to the manual recordings (for full description, see Supporting Information).

We measured the relative elevation of each local community to the nearest focal pond and calculated water depth as a mean daily value. From these data we calculated five different indices of water cover history for each plot (Table 1). The sampled abiotic parameters were condensed to a single primary axis of environmental variation using a Principal Components Analysis (PCA). The PCA yielded a PC1-axis accounting for 59% of the total variation and with similar loadings for most of the measured parameters (see Table S1). Both water cover parameters and soil/sediment depth decreased with increasing

PC1 value and the axis can largely be interpreted as a flooding gradient.

Functional plant traits Species’ traits were measured on individuals collected throughout their spatial distribution in the study site in May 2010 (with a few additional species collected in 2011 and 2012). Traits were measured on the 32 most dominant species of vascular plant species responsible for more than 80% of the relative abundance in the investigated plots.

We regarded the regional species pool as all the species occurring in the study area and for which the traits were known. The regional abundance could then be calculated as the relative number of subplots occupied by a given species. Because of the small area of the study site (~2 ha), dispersal limitation was considered to have negligible effects on the community assembly.

Because of our emphasis on environmental filtering acting on plant traits (see below), we focussed on functional traits relevant for the environmental gradient and measurable on a continuous scale.

Table 2. Covariance of traits within species assessed with Spearman rank correlation. In the upper right-hand triangle, p-values are shown, while r-values are in the lower left-hand triangle. A positive r-value indicates that species with high values for one trait also tends to have high values for the other.

Trait SLA LDMC LA Root porosity

WLOD Height

SLA (cm2 g-1) 0.14 0.48 0.39 0.0097** 0.28 LDMC (mg g-1) -0.27 0.68 0.50 0.18 0.70 Area cm2 -0.13 0.076 0.016* 0.063 0.047*Root porosity (%) 0.16 -0.12 0.42 0.27 0.22 WLOD (%) 0.45 0.24 -0.33 0.27 0.26 Height (m) -0.20 -0.072 0.35 0.22 -0.21

*p<0.05; **p<0.01

67

Community assembly from pond to alvar

Leaf area (cm2) was measured on the youngest yet fully developed and expanded leaf from at least 10 individual plants per species (excluding petiole if present) using at flatbed desktop scanner. The same leaves were weighed when fresh and fully hydrated, and subsequently after > 24 hours of drying at 105o C. Leaf dry matter content (LDMC, mg g-1) was assessed as leaf dry weight in relation to fresh weight. Specific leaf area (SLA, cm2 g-1) was calculated from leaf area and leaf dry weight. For one species, Schoenoplectus tabernaemontani (C.C.Gmel.) Palla, leaf area was estimated from area-length relations made from leaf sections and measured plant height. For a few species, less than 10 individuals were examined (see Table S2).

For the measurement of desiccation resistance and root porosity, plants were collected at the study site by digging up whole sods that were transported to the laboratory, placed in dim light at moderate air temperatures and analysed within one day of collection. Plant resistance to desiccation was measured as initial water loss on drying (WLOD %) at the observed maximum summer temperature. Three entire green shoots were separated from their roots at the soil surface, weighed and placed upright in a rack in an oven at 50o C for 20 minutes and reweighed. Drying was then continued for a minimum of 16 hours at 105o C to ensure loss of all remaining water and final weights were measured. WLOD was calculated as the % of the total water content lost during the initial 20 minutes at 50o C.

Root porosity was assessed using the gravitational method of Visser & Bögemann (2003). To ensure sufficient plant material (min. 100 mg per batch) for accurate determination, we often had to use more than one individual per analysis. The analysis were, except in a few instances perfomed in triplicates (Table S2). In short, roots were cleaned of all soil particles and debris and cut into ~2 cm long pieces. Fresh weight was recorded. The root fragments were then transferred to a small water-filled 20 ml glass

flask closed with a rubber stopper with an attached metal screw hook and then weighed submerged in water using a hanging balance. The flask was then transferred to a desiccator connected to a vacuum pump and subjected to 5 min of 0.8 bar partial vacuum three times in a row to replace the root gas with water. The flasks were then reweighed and the porosity calculated according to the formula:

where W1 is the weight of the flask with root material before vacuum minus the weight of the flask without roots and W2 is the weight of the flask with root material after vacuum minus the weight of the flask without roots. FW is the fresh weight of the used root material.

Plant height was extracted from the LEDA data base (Kleyer et al., 2008). Due to the much skewed distribution of plant height and leaf area these parameters were log transformed before analysis.

Environmental trait relationship Due to environmental filtering, the expression of functional plant traits by local communities are expected to vary along abiotic gradients (Shipley, 2010b). A community-weighted mean value (CWM) can be used as an estimate of the trait of an average individual in a community and is a suitable measure of the trait response along environmental gradients. It is calculated as:

where ti is the trait t of species i of S species in the local community each with a relative abundance pi. The relationship between the condensed environmental variables (PC1) and CWMs was determined by regression analysis, while the co-variation of traits within species was examined using Spearman rank correlation (only significant results are shown).

Predictability of the CWM values from environmental conditions is essential for the

,100(%)1

12

WFW

WWPorosity

,1

S

iii tpCWM

68

Community assembly from pond to alvar

implementation of the full Shipley model (Shipley et al., 2011) (see below). We quantified to what extent the environmental conditions can be used to accurately predict the CWM values observed in the local communities by generalized additive models (GAM) with CWM of each trait as response variable and the observed environmental conditions as predictor variables. Backward elimination was used for model simplification with the p = 0.05 level used to sequentially assess models.

To examine to what extent traits were subject to environmental filtering we used the approach of Ingram and Shurin (2009). Here the range of traits within a local community is compared to the range expected according to a null model derived from the regional species pool. In short, we produced the null model for every local community by simulating 999 communities with the same species number, but randomly including species from the regional species pool. The obtained null distribution of a trait range in a community was used to assess if the observed range within a community was smaller than expected for a randomly assembled community, i.e. clustered. For each trait in each community, we calculated the p value as the proportion of the 999 null communities with a larger trait range value than the observed. If the range was smaller than 97.5% it was accepted as being significant with =0.05 (Ingram and Shurin 2009).

To alleviate multiple comparisons (6 traits in 21 communities) we also conducted the ‘meta-analysis’ proposed by Ingram and Shurin (2009) by first calculating the Trait Clustering Index (TCI) for each community as:

where Rangeobs is the observed trait range,

is the average range of the null

model and SD(Rangenull) is the standard deviation of the null model. Thus positive values of TCI indicate lower range of a given

trait than expected by random sampling. We then tested whether the TCI values of all the analyzed communities deviated from zero using a one sample t-test to indicate overall tendencies towards trait clustering. Both the GAMs and the trait clustering analysis were performed in the R environment (R development Core Team, 2012) using mgcv library and the library provided by Ingram & Shurin (2009) respectively.

Trait abundance models Shipley et al. (2006) developed a model, using maximum entropy with community aggregated traits as constraints to predict the relative abundance of species within a local community from a regional species pool. We used this maxent (maximum entropy) based model to predict the relative abundances of species in the studied communities from the measured traits. We tested the model with permutation tests (Shipley, 2010c) and used the backward stepwise procedure proposed by Shipley et al. (2011) to create the simplest model that, given the measured traits, still yielded a result significantly better than random. Earlier work has shown that using more informative priors improve the models predictive ability (Sonnier et al., 2010, Laliberté et al., 2012). Therefore we ran the model using the regional relative abundance of each species as priors. The resulting model did not have markedly improved prediction capacity, and we thus only present results from the model using the maximally uninformative prior. The maxent model estimates a series of constants (λ) (Lagrange multipliers) quantifying how much a unit change in a particular trait is associated with a proportional change in the predicted relative abundance in a local community when all other traits are kept constant. Hence, λ assesses the magnitude with which the relative abundance of a species changes as the value of the focal trait changes. Positive λ-values mean that locally abundant species have large values for the focal trait.

,

nullRange

69

Community assembly from pond to alvar

Fig 1. Community-weighted mean trait values (CWM, black dots) and ranges of trait values (bars) in local communities along a hydrological gradient (expressed as PC1) of increasingly drier conditions. The curves represent least square regression models and r2-values indicate strength of model fit. Linear regression were used for leaf dry matter content (LDMC) while second order polynomial regression lines were fitted to specific leaf area (SLA), water loss on drying (WLOD), plant height, leaf area (log), and root porosity Conversely, a negative λ means that species with large values for the focal trait are subordinate or absent. Values of λ close to zero mean that there is no directional selection of the particular trait in the local community (Shipley, 2010b). To ensure comparability between the λ –values all traits were standardized to unit variance before analysis. All models were performed in the R environment using the FD library (Laliberté & Shipley, 2009, R development Core Team, 2012).

Results Environment- trait relationships The environmental conditions in local communities were highly variable with some plots being permanently submerged with quite variable average sediment depth (2.3-16.1 cm) while other plots were never flooded and generally had very shallow soil layers (0.7-5.2 cm, Table 1).

Across species, some traits showed indications of covariance. Species with large

0

200

400

600r2= 0.44

SLA

(cm

2 g

-1)

-6 -4 -2 0 2 40

10

20

30

40r2= 0.59

PC1

Roo

t por

osity

(%

)

0

100

200

300

400 r2= 0.31

LDM

C (

mg

dw/g

fw

)

0

20

40

60

80

100r2= 0.41

WLO

D (

%)

-6 -4 -2 0 2 4-2

-1

0

1

2

3r2= 0.65

PC1

log

(leaf

are

a, c

m2 )

0.0

0.5

1.0

1.5

2.0r2= 0.55

Hei

ght (

m)

70

Community assembly from pond to alvar

leaf area also tended to be tall and have high root porosity, while species with high SLA rapidly lost a large percentage of their water content upon desiccation at 50 o C (WLOD, Table 2).

The community-weighted mean values (CWM) responded to the environmental gradient as indicated by significant regressions (Fig. 1). Leaf dry matter content (LDMC) responded linearly to the gradient while the response of the remaining traits was significantly better described by a quadratic function indicating that most traits respond in a nonlinear fashion to the investigated gradient. Significant GAM models with high predictive ability revealed strong relationships between environmental conditions and CWM trait values (Table 3). All analysed environmental variables contributed significantly to one or more of the models. Variables such as summed water depth and vegetation height contributed to the majority of the models and only the model of leaf area did not include any of the hydrological variables (Table 3).

Table 3. Generalized additive models (GAM) of the community-weighted mean trait values (CWM) using the 14 environmental variates as predictor terms (as numbered in Table 1). CWM Trait Environmental variate no. R2 SLA 4, 5, 3, 2, 7, 8, 6, 9, 11 0.87 LDMC 1, 4, 5, 14 0.81 LA 8, 9, 10, 12 0.83 Root porosity 3, 2, 6, 9, 10, 12 0.82 WLOD 4, 5, 3, 8, 6, 9, 10, 11 0.80 Height 4, 7, 8, 6, 9, 10, 11, 12, 13 0.92

Displaying the CWM and range of traits along the condensed environmental gradient (PC1) revealed that leaf area and plant height decreased, while LDMC increased, along the gradient, which means that shoot and leaf size decreased towards dryer conditions, while leaves contained more dry mass per fresh weight (Fig 1). Root porosity appeared to show a saturation response being low on dry and shallow soils, while increasing under wetter conditions and saturating at 30% aerenchyma under permanently submerged conditions. SLA and WLOH peaked at intermediate flooding and soil depths (Fig. 1).

Table 4. Trait filtering and its importance for community assembly along a hydrological gradient. Environmental filtering was evaluated using the trait clustering index (TCI), where values above zero indicate a lower range of trait values within a community than expected from a null model. Significance was tested as deviation from zero by a one sample t-test. Trait importance to community assembly was assessed by the absolute λ-values from a maxent model containing all traits. Consistent change of the degree of filtering and importance along the gradient was assessed by correlation of TCI and λ-values, repectively, with the condensed environmental gradient (PC1). A positive r-value indicates higher trait filtering in the part of gradient with drier conditions and shallower soils, while negative values indicate stronger filtering under conditions with frequent flooding or stagnant water.

Trait Average TCI TCI : PC1 Spearman r

Absolute λ-values

λ : PC1 Spearman r

Specific leaf area 0.55 ** 0.62 ** 33.6 0.07 Water loss on drying 0.52 * 0.26 41.8 -0.34 Root porosity 0.81 * -0.58 ** 30.5 -0.44 * Leaf dry matter content 0.45 -0.86 ** 25.9 0.61 ** Plant height† 0.27 0.70 ** 18.0 -0.12 Leaf Area† 0.58 0.63 ** 27.2 -0.32

† Variable was log transformed before analysis *p<0.05; **p<0.01

71

Community assembly from pond to alvar

Trait filtering across the gradient In up to seven of the 21 local communities, the observed range of a given trait was significantly smaller than expected from the null models. Within each local community between zero and three traits showed significant clustering (Table S3).

A large standardized trait clustering metric (TCI) causing a significant deviation from zero across communities provides evidence of environmental filtering along the gradient. We found significant evidence for environmental filtering acting through the plant traits root porosity, WLOD, SLA, and a tendency for leaf area to experience filtering as well (Table 4).

SLA and WLOD showed substantial co-variation (Table 2), with species having a high SLA also exhibiting a high WLOD. The two factors are probably significant for the same reason – variation in resistance to water loss. Root porosity - reflecting the ability to supply oxygen to root meristems under wet, anoxic conditions in soils - was not correlated to either SLA or WLOH. Hence, we found evidence for environmental filtering in two independent traits – tolerance to drought versus to flooded soils along the sampled gradient.

Fig. 2. Trait clutering index (TCI) in the local communities along the hydrological gradient (expressed as PC1). Values of TCI above zero indicate environmental filtering of species through the measured traits. The TCI values of SLA, WLOD and root porosity are significantly different from zero across the gradient (one-sided t-test, p<0.05)

-6 -4 -2 0 2 4

-2

0

2

4r=-0.85*

TCI L

DM

C

-6 -4 -2 0 2 4

-2

0

2

4r=0.25

TCI W

LO

D

-6 -4 -2 0 2 4

-2

0

2

4r=0.68*

TCI h

eig

ht

-6 -4 -2 0 2 4

-2

0

2

4r=0.63*

TCI S

LA

-6 -4 -2 0 2 4

-2

0

2

4r=0.62*

PC1

TCI L

A

-6 -4 -2 0 2 4

-2

0

2

4r=-0.59*

PC1

TCI p

oro

sity

72

Community assembly from pond to alvar

Displaying the TCI values along the sampled gradient (PC1) allows for an assessment of the degree of environmental filtering of a given trait under different conditions (Fig. 2, Table 4). For all traits, except WLOD, the degree of trait clustering is correlated significantly to the overall gradient (Table 3) showing that although a given trait is not filtered by the environment under all the sampled conditions, it might still be strongly range limited along parts of the gradient. Prediction of community structure From the observed CWM trait values and the individual trait values of each species in the species pool, we predicted the relative abundance of each species in local communities using the maxent model. The model including all six traits predicted 66 % (p = 0.03) of the observed relative abundances of the 32 species included in the species pool (Fig. 3). The backward elimination procedure resulted in a significant model containing only one trait (leaf area, p= 0.02) and, with a substantially lower predictive ability (r2 = 0.10) (Fig. 4).

Fig. 3. Predictions of relative abundance of plant species from the maxent model vs. observed relative abundances. The maxent model was based on all measured traits, the regional species pool and the 21 local communities. The line indicates the y=x relationship.

From the maxent model using all the measured traits as constraints, the λ values were obtained for each trait in each local community (see Fig. S5). The λ values varied greatly between the communities with both positive and negative values indicating that selective pressure on traits differs within short spatial distances as a result of the steeply changing environmental conditions. The change in λ values along the gradient was consistent for root porosity and LDMC, as they all correlated significantly with the environmental gradient (PC1-axis, Table 4). λporosity changed from positive values to values close to zero, showing that high root porosity is a necessity for growth and survival under waterlogged conditions, while the trait decreased in importance under drier conditions, possibly even becoming disadvantageous in the driest sites. The opposite trend was found for λLDMC showing that a large dry mass per fresh mass was undesirable in wet sites, but seemingly unimportant under drier conditions. The λ-values for the remaining traits showed large variations and did not change consistently along the gradient and thus do not appear to play any consistent role along the investigated hydrological gradient.

Overall, using only six traits in the model predicted species’ relative abundance fairly well and root porosity along with LDMC appeared to play the most consistent role for community assembly along the gradient.

Discussion Relationship between environment and traits The environmental conditions at the study site range from very harsh - with plants growing in soils less than one cm deep and often experiencing summer temperatures above 40oC - through more productive areas with deeper soil, higher availability of soil nutrients and water - to areas with a permanent water cover. Water cover reduces inorganic carbon supply to most plants and generates anoxic sediments. Traits known to be crucial to plant survival and growth under

0.0 0.5 1.00.0

0.5

1.0

r2=0.66

Observed abundance

Pre

dict

ed a

bund

ance

73

Community assembly from pond to alvar

these contrasted and adverse conditions, such as root porosity under flooding (Justin & Armstrong, 1987, Colmer & Voesenek, 2009) and leaf dry matter content (LDMC) under drought (Hallik et al., 2009), changed accordingly. Overall, we found that community-weighted mean values (CWM) of measured traits were strongly related to the position along the environmental gradient of the particular community.

Using generalized additive models (GAM), we found a close relationship between CWM trait values and environmental conditions. In particular, hydrological properties associated with water depth and frequency of flooding were important predictors for most CWM trait values. Vegetation height, as a proxy for site productivity, also contributed significantly to models for several traits. This property could probably be better substituted by direct measurements of available soil nutrients (Pérez-Ramos et al., 2012).

The strong correlation of SLA with resistance to desiccation (WLOD) underpins the general importance of this often-measured trait (Wright et al., 2004). Unexpectedly, however, SLA and WLOD showed unimodal responses to the hydrological gradient. In the drier sites, SLA increased with soil depth, which corresponds well with previous findings (Bernard-Verdier 2012). The decrease of SLA with depth and duration of water cover may seem counterintuitive and is at odds with previous findings (Violle et al., 2011). However, there is probably a fundamental difference in the function of SLA between flooded meadows and permanent water with deeper sediments. At the study site, low-SLA species are mainly Typha latifolia L and Potamogeton natans L. For the emergent macrophyte T. latifolia, characteristic of environments with ample nutrients and (excess) water (Wisheu & Keddy, 1992), low SLA may be a trait related to competition attained through accumulation of a recalcitrant litter layer (Christensen et al., 2009).

Fig. 4. Ability of the maxent model to predict the relative abundance of the species in the investigated communities (Pearsons r2). The sequence of traits is based on the backward elimination procedure to obtain the simplest model, starting with the full model with all six traits included in the right side of the graph. The traits were eliminated based on the size of their absolute λ values. All models were significantly better than random (p<0.05)

For the floating-leaved Potamogeton

natans, low SLA may express resistance to high mechanical stress from wave action (Puijalon et al., 2011). For permanently submerged species, SLA would indeed be high (Nielsen & Sand-Jensen, 1989)

Environmental filtering and trait clustering Across the environmental gradient three traits were significantly clustered, indicating that environmental filtering plays a major role for community assembly under the highly variable conditions investigated.

All traits showed some degree of filtering and in five of six analyzed traits, the degree of filtering correlated with the hydrological gradient. Flooding and drought have been known since long to strongly affect trait composition in plant communities (Iversen, 1936, van Eck et al., 2004, Jacobsen et al., 2008). SLA and WLOD seemed to be environmentally filtered across the entire gradient, which corresponds with findings for both woody (Cornwell & Ackerly, 2009) and perennial herbaceous species (Bernard-Verdier et al., 2012).

0 2 4 60.0

0.2

0.4

0.6

0.8

LA

Porosity

WLOD

SLA

LDMC

Height

Traits

r2

74

Community assembly from pond to alvar

Fig. 5. Relationship between the trait clustering index (TCI) and absolute λ-values in the 21 local communities. TCI is based on the range of traits observed in a given community compared to a null model, while the λ-value indicates the importance of the trait for prediction in the maxent model. Positive correlation between the two parameters indicates that in communities where a given trait experiences strong filtering the same trait also plays an important role in community assembly

In contrast to the mentioned studies, we

found a relationship between the degree of filtering in SLA and the hydrological gradient, indicating trait clustering on shallower soils with drier conditions. Most likely, very dry sites demand special strategies for leaf water economy, whereas

pond borders and submerged sites offer more room for trait variation. Not considering submerged communities with helophytes only, WLOD showed the same pattern as SLA with drier sites having stronger filtering than wetter. Leaf area and plant height, although apparently not environmentally

-2 0 2

0

5

10

r=0.64*

TCILDMC

LD

MC

-2 0 2

0

5

10

15r=-0.16

TCIWLOD

WL

OD

-2 0 2

0

5

10r=-0.07*

TCILA

LA

-2 0 2

0

5

10

15r=-0.44*

TCISLA

SL

A

-2 0 2

0

5

10r=-0.026

TCIHeight

Hei

ght

-2 0 2

0

5

10r=0.48*

TCIporosity

po

rosi

ty

75

Community assembly from pond to alvar

filtered along the whole gradient, both showed signs of filtering towards the dry end of the gradient, most likely caused by the same factors as described for SLA and WLOD.

Root porosity was subjected to environmental filtering across the hydrological gradient. This is, to our knowledge, the first time that root porosity has been shown to undergo environmental filtering. The trait range was particularly restricted – all species having high root porosity - under submerged conditions, as expected from the function of aerenchyma in root oxygen supply in anoxic sediments (Justin & Armstrong, 1987, Douma et al., 2012). The filtering of root porosity also in some communities at the dry end of the gradient could by caused by a tradeoff between root porosity and nutrient uptake or mechanical strength (Končalová, 1990, Striker et al., 2007).

The maxent model presumes that environmental filtering is important to community structure (Shipley, 2010a, Shipley, 2010b). One should therefore anticipate that traits under strong environmental filtering (i.e. high TCI) would also be the best predictors of community assembly according to the maxent model (large absolute λ values). The three traits that experienced significant environmental filtering (SLA, WLOD and root porosity) across communities also appeared to be the better predictors of community assembly, as these traits had the highest λ-values in the full maxent model. The same three traits, along with leaf area also showing indications of filtering across the gradient, remained in the model simplified by backward elimination suggesting a consistent importance for community assembly. The seeming importance of leaf area was mainly driven by one of the investigated 21 local communities, perhaps putting a question mark to the use of backward elimination for finding the most parsimonious version of the maxent model. Including more community samples could

potentially alleviate this effect, as single λ values would have smaller impact on the summed λ used as selection criteria.

Similar to values of the Trait Clustering Index, λ values for local communities taken from the full model varied considerably along the environmental gradient, an observation also made elsewhere (Laliberté et al., 2012, Sonnier et al., 2012). The fact that metrics of both methods from the single local communities correlated with the environmental gradient called for an assessment of TCI and λ values on a local community scale (Fig. 5). For root porosity and LDMC, a significant positive correlation between TCI and the absolute λ values indicated that traits experiencing strong filtering (higher TCI values) are also strong determinants of community assembly (Fig. 5). For the remaining traits and plant height, the relationship between trait filtering and trait importance was either not significant or negative because the λ-values did not respond monotonously along the environmental gradient. Thus, the correspondence between the independent measures of environmental filtering and traits important to community assembly confirms that environmental filtering acts through traits that have great influence on community assembly. Nevertheless, the lack of complete congruence of the two methods may result from the fact that the maxent model does not fully explain the observed relative abundance (r2 = 0.66). This in turn could be a result of not including all traits of importance for community assembly (Sonnier et al., 2010). Traits that could have proven valuable for community assembly are seed germination above and under water (van der Valk, 1981) and seed buoyancy (Douma et al., 2012), that previously have been shown to be important for plant distribution along flooding gradients.

In conclusion, we found that both community-weighted mean traits and environmental filtering of functional traits of species responded to a gradient in water coverage and soil depth. Using the maxent

76

Community assembly from pond to alvar

approach, we could make powerful predictions of relative abundance of plant species using only six measured traits. We found evidence for congruence between the degree of filtering experienced by a trait and its importance for predicting community assembly.

Acknowledgements L.B.S. was funded by a PhD-grant from the Faculty of Science, University of Copenhagen. This work was supported by Center of lake restoration (CLEAR) a Villum Kann Rasmussen Center of Excellence to K.S.J. and a grant from the Carlsberg foundation to K.S.J. (Ecology of Nostoc).

References Ackerly, D. D. & Cornwell, W. K. (2007) A trait-

based approach to community assembly: partitioning of species trait values into within- and among-community components. Ecology Letters, 10, 135-145.

Bernard-Verdier, M., Navas, M.-L., Vellend, M., Violle, C., Fayolle, A. & Garnier, E. (2012) Community assembly along a soil depth gradient: contrasting patterns of plant trait convergence and divergence in a Mediterranean rangeland. Journal of Ecology, 100, 1422-1433.

Christensen, J. P. A., Sand-Jensen, K. & Staehr, P. A. (2013) Fluctuating water levels control water chemistry and metabolism of a charophyte dominated pond. Freshwater Biology, DOI: 10.1111/fwb.12132.

Christensen, J. R., Crumpton, W. G. & van der Valk, A. G. (2009) Estimating the breakdown and accumulation of emergent macrophyte litter: A mass-balance approach. Wetlands, 29, 204-214.

Colmer, T. D. (2003) Long-distance transport of gases in plants: a perspective on internal aeration and radial oxygen loss from roots. Plant Cell and Environment, 26, 17-36.

Colmer, T. D. & Voesenek, L. (2009) Flooding tolerance: suites of plant traits in variable environments. Functional Plant Biology, 36, 665-681.

Cornwell, W. K. & Ackerly, D. D. (2009) Community assembly and shifts in plant trait

distributions across an environmental gradient in coastal California. Ecological Monographs, 79, 109-126.

Crawford, R. M. M. (2009) Survival at the water’s edge. Plants at the Margin: Ecological Limits and Climate Change. Cambridge University Press, Cambridge.

Douma, J. C., Bardin, V., Bartholomeus, R. P. & van Bodegom, P. M. (2012) Quantifying the functional responses of vegetation to drought and oxygen stress in temperate ecosystems. Functional Ecology, 26, 1355-1365.

Grime, J. P. (1998) Benefits of plant diversity to ecosystems: immediate, filter and founder effects. Journal of Ecology, 86, 902-910.

Grime, J. P. (2006) Trait convergence and trait divergence in herbaceous plant communities: Mechanisms and consequences. Journal of Vegetation Science, 17, 255-260.

Hallik, L., Niinemets, U. & Wright, I. J. (2009) Are species shade and drought tolerance reflected in leaf-level structural and functional differentiation in Northern Hemisphere temperate woody flora? New Phytologist, 184, 257-274.

Hooper, D. U., Chapin, F. S., Ewel, J. J., Hector, A., Inchausti, P., Lavorel, S., Lawton, J. H., Lodge, D. M., Loreau, M., Naeem, S., Schmid, B., Setälä, H., Symstad, A. J., Vandermeer, J. & Wardle, D. A. (2005) Effects of biodiversity on ecosystem functioning: A consensus of current knowledge. Ecological Monographs, 75, 3-35.

Ingram, T. & Shurin, J. B. (2009) Trait-based assembly and phylogenetic structure in northeast Pacific rockfish assemblages. Ecology, 90, 2444-2453.

Iversen, J. (1936) Biologische Pflanzentypen als Hilfsmittel in der Vegetationsforschung. Ein Beitrag zur ökologischer Charakterisierung und Anordnung der Pflanzengesellshaften. Levin & Munksgaard, Copenhagen.

Jacobsen, A. L., Pratt, R. B., Davis, S. D. & Ewers, F. W. (2008) Comparative community physiology: nonconvergence in water relations among three semi-arid shrub communities. New Phytologist, 180, 100-113.

Justin, S. H. F. W. & Armstrong, W. (1987) The anatomical characteristics of roots and plant response to soil flooding. New Phytologist, 106, 465-495.

77

Community assembly from pond to alvar

Keddy, P. A. (1992) Assembly and response rules - two goals for predictive community ecology. Journal of Vegetation Science, 3, 157-164.

Kleyer, M., Bekker, R. M., Knevel, I. C., Bakker, J. P., Thompson, K., Sonnenschein, M., Poschlod, P., van Groenendael, J. M., Klimeš, L., Klimešová, J., Klotz, S., Rusch, G. M., Hermy, M., Adriaens, D., Boedeltje, G., Bossuyt, B., Dannemann, A., Endels, P., Götzenberger, L., Hodgson, J. G., Jackel, A. K., Kühn, I., Kunzmann, D., Ozinga, W. A., Römermann, C., Stadler, M., Schlegelmilch, J., Steendam, H. J., Tackenberg, O., Wilmann, B., Cornelissen, J. H. C., Eriksson, O., Garnier, E. & Peco, B. (2008) The LEDA Traitbase: a database of life-history traits of the Northwest European flora. Journal of Ecology, 96, 1266-1274.

Končalová, H. (1990) Anatomical adaptations to waterlogging in roots of wetland graminoids: Limitations and drawbacks. Aquatic Botany, 38, 127-134.

Laliberté, E. & Shipley, B. (2009) FD: Measuring the functional diversity from multiple traits, and other tools for functional ecology. pp. R package. The comprehensive R archive Network (CRAN), Vienna, Austria.

Laliberté, E., Shipley, B., Norton, D. A. & Scott, D. (2012) Which plant traits determine abundance under long-term shifts in soil resource availability and grazing intensity? Journal of Ecology, 100, 662-677.

Lavorel, S. (2013) Plant functional effects on ecosystem services. Journal of Ecology, 101, 4-8.

Madsen, T. V. & Sand-Jensen, K. (1991) Photosynthetic carbon assimilation in aquatic macrophytes. Aquatic Botany, 41, 5-40.

McGill, B. J., Enquist, B. J., Weiher, E. & Westoby, M. (2006) Rebuilding community ecology from functional traits. Trends in Ecology & Evolution, 21, 178-185.

Mommer, L., Lenssen, J. P. M., Huber, H., Visser, E. J. W. & de Kroon, H. (2006) Ecophysiological determinants of plant performance under flooding: a comparative study of seven plant families. Journal of Ecology, 94, 1117-1129.

Mommer, L. & Visser, E. J. W. (2005) Underwater photosynthesis in flooded terrestrial plants: A matter of leaf plasticity. Annals of Botany, 96, 581-589.

Nielsen, S. L. & Sand-Jensen, K. (1989) Regulation of photosynthetic rates of submerged rooted macrophytes. Oecologia, 81, 364-368.

Pérez-Ramos, I. M., Roumet, C., Cruz, P., Blanchard, A., Autran, P. & Garnier, E. (2012) Evidence for a 'plant community economics spectrum' driven by nutrient and water limitations in a Mediterranean rangeland of southern France. Journal of Ecology, 100, 1315-1327.

Puijalon, S., Bouma, T. J., Douady, C. J., van Groenendael, J., Anten, N. P. R., Martel, E. & Bornette, G. (2011) Plant resistance to mechanical stress: evidence of an avoidance-tolerance trade-off. New Phytologist, 191, 1141-1149.

R development Core Team (2012) R: a language and environment for statistical computing. R foundation for statistical computing, Vienna, Austria.

Sand-Jensen, K., Baastrup-Spohr, L., Møller, C. L., Winkel, A., Brodersen, K. P., Borum, J. & Staehr, P. A. (2010) Plant distribution patterns and adaptation in a limestone quarry on Öland. Svensk Botanisk Tidsskrift, 104, 23-31.

Sand-Jensen, K. & Jespersen, T. S. (2012) Tolerance of the widespread cyanobacterium Nostoc commune to extreme temperature variations (-269 to 105 degrees C), pH and salt stress. Oecologia, 169, 331-339.

Sand-Jensen, K. & Staehr, P. A. (2012) CO2 dynamics along Danish lowland streams: water–air gradients, piston velocities and evasion rates. Biogeochemistry, 111, 615-628.

Shipley, B. (2010a) Community assembly, natural selection and maximum entropy models. Oikos, 119, 604-609.

Shipley, B. (2010b) From plant traits to vegetation structure - Chance and selection in the assembly of ecological communities. Cambridge University Press, Cambridge.

Shipley, B. (2010c) Inferential permutation tests for maximum entropy models in ecology. Ecology, 91, 2794-2805.

Shipley, B., Laughlin, D. C., Sonnier, G. & Otfinowski, R. (2011) A strong test of a maximum entropy model of trait-based community assembly. Ecology, 92, 507-517.

Shipley, B., Vile, D. & Garnier, E. (2006) From plant traits to plant communities: A statistical mechanistic approach to biodiversity. Science, 314, 812-814.

78

Community assembly from pond to alvar

Silvertown, J., Dodd, M. E., Gowing, D. J. G. & Mountford, J. O. (1999) Hydrologically defined niches reveal a basis for species richness in plant communities. Nature, 400, 61-63.

SMHI (2013) Swedish Meterological and Hydrological Institue, http://www.smhi.se/klimatdata/meteorologi/temperatur/dataserier-med-normalv%C3%A4rden-1.7354.

Sonnier, G., Navas, M. L., Fayolle, A. & Shipley, B. (2012) Quantifying trait selection driving community assembly: a test in herbaceous plant communities under contrasted land use regimes. Oikos, 121, 1103-1111.

Sonnier, G., Shipley, B. & Navas, M. L. (2010) Plant traits, species pools and the prediction of relative abundance in plant communities: a maximum entropy approach. Journal of Vegetation Science, 21, 318-331.

Striker, G. G., Insausti, P., Grimoldi, A. A. & Vega, A. S. (2007) Trade-off between root porosity and mechanical strength in species with different types of aerenchyma. Plant Cell and Environment, 30, 580-589.

van der Valk, A. G. (1981) Succession in wetlands: A Gleasonian approach. Ecology, 62, 688-696.

van Eck, W., van de Steeg, H. M., Blom, C. & de Kroon, H. (2004) Is tolerance to summer flooding correlated with distribution patterns in river floodplains? A comparative study of 20 terrestrial grassland species. Oikos, 107, 393-405.

Violle, C., Bonis, A., Plantegenest, M., Cudennec, C., Damgaard, C., Marion, B., Le Cæur, D. & Bouzillé, J. B. (2011) Plant functional traits

capture species richness variations along a flooding gradient. Oikos, 120, 389-398.

Visser, E. J. W. & Bögemann, G. M. (2003) Measurement of porosity in very small samples of plant tissue. Plant and Soil, 253, 81-90.

Warming, E. (1895) Plantesamfund - Grundtræk af den økologiske plantegeografi. P. G. Phillipsen, Copenhagen.

Wetzel, R. G. (2001) Limnology: Lake and River Ecosystems. Elsevier, San Diego, California.

Wisheu, I. C. & Keddy, P. A. (1992) Competition and centrifugal organization of plant communities: theory and tests. Journal of Vegetation Science, 3, 147-156.

Wright, I. J., Reich, P. B., Westoby, M., Ackerly, D. D., Baruch, Z., Bongers, F., Cavender-Bares, J., Chapin, T., Cornelissen, J. H. C., Diemer, M., Flexas, J., Garnier, E., Groom, P. K., Gulias, J., Hikosaka, K., Lamont, B. B., Lee, T., Lee, W., Lusk, C., Midgley, J. J., Navas, M. L., Niinemets, U., Oleksyn, J., Osada, N., Poorter, H., Poot, P., Prior, L., Pyankov, V. I., Roumet, C., Thomas, S. C., Tjoelker, M. G., Veneklaas, E. J. & Villar, R. (2004) The worldwide leaf economics spectrum. Nature, 428, 821-827.

Supplementary material The supplementary material are palced at then end of the thesis. It consists of a thorough explanation of water indices, PCA-tables, tables of all measured traits, tables of TCI-values for all trait in all communities and a graphs showing the relationships between λ-values and the PCA-axis.

79

 

80

81

 

82

Isoedtids and water level fluctuations  

Water level fluctuations affect plant growth and sediment dynamics of isoetid populations in oligotrophic lakes Lars Baastrup-Spohr*, Claus Lindskov Møller and Kaj Sand-Jensen Freshwater Biological Laboratory, Biological Institute, University of Copenhagen, Universitetsparken 4, 3. Floor, 2100 Copenhagen, Denmark

1. Several amphibious isoetid species, and Littorella uniflora in particular, exchange O2 and CO2 across roots with the sediment and dominate the littoral zone from below to above the water in North temperate oligotrophic, softwater lakes. Accumulation of labile organic matter in the sediments from phytoplankton blooms accompanying eutrophication represents a threat to isoetids because of insufficient root oxygenation, reduced anchorage and enhanced growth of taller competitors. We hypothesized that L. uniflora tolerates different water levels but grows better and maintains dominance when sediments are periodically drained because air exposure assists oxygenation, mineralization and consolidation of sediments. If so, drainage offers a possible restoration tool. We tested this hypothesis by measuring plant performance and sediment metabolism under drained, fluctuating and submerged conditions in the laboratory and along elevation gradients in the field in natural and organically enriched sandy sediments from an oligotrophic Danish lake. 2. Laboratory and field experiments confirmed that L. uniflora maintained high leaf chlorophyll and photosynthesis and grew in biomass under all conditions. Increase of biomass and shoot weight was highest under drained conditions and most so for roots as a suitable response to higher sediment oxygenation and leaf transpiration. The increase of biomass was lowest under fluctuating conditions as a logical consequence of higher costs of frequently producing aerial leaves to replace unsuitable aquatic leaves upon changes from submerged to drained conditions. 3. Efflux of inorganic carbon from mineralization in un-vegetated standard sediments increased 5.5-fold from submerged, over fluctuating to drained treatments. In submerged vegetated sediments, where part of sediment CO2 is incorporated into a growing plant biomass, the sum of inorganic carbon efflux and plant growth had more than doubled compared to bare sediments probably reflecting stimulated sediment oxygenation and mineralization because of root oxygen release. Under drained, well-oxygenated conditions the increase of carbon in the biomass was accompanied by an equivalent decline of the inorganic carbon efflux relative to bare sediments. Still, the sum of inorganic carbon efflux and additional carbon trapped in the biomass was 2-fold higher under drained than submerged conditions. 4. The proportion of reduced Fe2+ to total Fe as a measure of the oxygen status of sediments declined from submerged, over fluctuating to permanently drained sediment. Also the proportion of reduced Fe2+ was lower in plant covered than bare sediments reflecting root release of O2. Drainage and air contact also increased sediment density and consolidation and, thereby, improved root anchorage compared to submerged conditions. 5. Periodic drainage and air exposure of isoetid covered sediments is a natural phenomenon in oligotrophic seepage lakes that should be reinstalled as a possible restoration tool if the water has been regulated to a constant level. Water level fluctuations of a moderate frequency can benefit isoetid populations because they grow and survive better than competitors and potential harmful organic matter can be mineralized without sediment anoxia.

Keywords: Softwater lake, isoetid, restoration, Litorella uniflora, water level fluctuation, sediment desiccation.

83

Isoedtids and water level fluctuations  

Introduction

The shallow littoral zone along rivers and streams and at the margins of lakes and ponds are often characterized by greatly fluctuating water levels from winter to summer, from year to year and over decades (van der Valk, 2005, White et al., 2008) influencing community assembly and growth of plant species (Baastrup-Spohr et al., 2013) and sediment processes (Baldwin & Mitchell, 2000). Fluctuations between submergence and emergence alter the availability of O2 and CO2 to photosynthesis and respiration because of 104-fold lower gas diffusion coefficients in water than in air. These large differences in abiotic conditions have profound effects on plant performance along hydrological gradients from permanently dry to permanently submerged conditions and are, thus, main determinants for the distribution of plant species along the margins of water bodies (van der Valk, 1981, Keddy & Reznicek, 1986, Baastrup-Spohr et al., 2013). In natural lakes, plant species richness have been shown to be positively related to the magnitude of water level fluctuations (Mjelde, Hellsten & Ecke, 2013) while the presence of single species are also highly dependent on the timing and frequency of water level fluctuations determining which species can germinate and thrive (van der Valk, 1981, Riis & Hawes, 2002, Van Geestet al., 2005).

In the zone of fluctuating water levels, amphibious plants often dominate the vegetation because they are capable of photosynthesizing and growing both under water and in air (Pedersen & Sand-Jensen, 1997, Casanova & Brock, 2000). They can accomplish this double-life by forming rigid “aerial leaves” with stomata and a well-developed cuticle to reduce evaporation in air and having flexible, thinner gas-permeable “aquatic leaves” without stomata under water (Nielsen, 1993, Robe & Griffiths, 1998, Mommer & Visser, 2005). Some amphibious species shift between utilizing atmospheric

CO2 by aerial leaves and combing passive CO2 use with active use of dissolved HCO3

- by aquatic leaves to compensate for lower gas diffusion rates in water (Sand-Jensen, Pedersen & Nielsen, 1992, Sand-Jensen & Frost-Christensen, 1999, Maberly & Madsen, 2002). Most amphibious leaves, however, rely on passive CO2 use from either air, water or, in some cases, the rhizosphere of alternating importance in response to fluctuating water levels and CO2 concentrations (Madsen & Sand-Jensen, 1991, Nielsen, Gacia & Sand-Jensen, 1991).

Flooded soils is another strong selective force which requires development of aerenchyma to ensure a sufficient O2 supply from leaves and stems to apical root meristems buried in anoxic soils (Justin & Armstrong, 1987, Colmer, 2003). Exposure to air for submerged plants, in contrast, represents a strong desiccation risk although some leaves may survive in the damp layer at the sediment surface (Sculthorpe, 1967) or rhizomes and roots may survive for a short while in wet soils without the supply of leaf photosynthates (Engelhardt, 2006). Also, terrestrial species can survive flooding and aquatic species survive drought as resistant seeds and sprout and grow when conditions once again becomes suitable to them (Brock et al., 2003). Thus, the zone of fluctuating water levels is inhabited by a mixture of terrestrial, amphibious and submerged species, although the amphibious species in terms of biomass and cover is the main vegetation component, while many terrestrial species may occur in low numbers scattered throughout the zone (Keddy & Reznicek, 1986, Casanova & Brock, 2000).

The sandy shores and shallow waters with fluctuating water levels in oligotrophic, softwater seepage lakes in the North temperate regions are often inhabited by small isoetid species of the rosette growth form (e.g., Carex viridula var. viridula, Elatine hexandra, Juncus bulbosus, Littorella uniflora, Lobelia dortmanna and Isoetes spp.; (Sand-Jensen & Søndergaard, 1979,

84

Isoedtids and water level fluctuations  

Smolders, Lucassen & Roelofs, 2002). In this zone of fluctuating water levels, the amphibious isoetid species can benefit from the reduced competition from taller, more nutrient demanding and less disturbance tolerant terrestrial and aquatic plants (Farmer & Spence, 1986). In NW Europe, Litorella uniflora is usually the dominant species in this habitat (Arts & Denhartog, 1990, Riis & Sand-Jensen, 1998, Pedersen et al., 2006) and it can form almost monospecific stands from 1 m above to 2-3 m below the mean water level (Sand-Jensen & Søndergaard, 1979). With 2-3-years intervals, minimum summer water level can change by up to 0.5-1.0 m between particularly wet and dry years (Sand-Jensen and Søndergaard 1979). Littorella uniflora is apparently well-adapted to fluctuating water levels by forming aerial leaves with stomata and low evaporation and submerged, gas-permeable and thicker leaves without stomata (Hostrup & Wiegleb, 1991, Robe & Griffiths, 1998). Leaves, stems and roots have well-developed air lacunae and the CO2-rich rhizosphere represents an important CO2 source to photosynthesis both in air and, even more so, under water (Nielsen et al., 1991).

Because of the highly permeable roots allowing uptake of CO2, Littorella uniflora and other isoetids are highly susceptible to low oxygen concentrations in the sediment often accompanying eutrophication, phytoplankton blooms and associated accumulation of labile organic matter in sediments (Møller & Sand-Jensen, 2011). Accumulation of organic matter not only affects the roots directly but can also lead to poorer root anchorage in looser sediments of higher water content leading to higher risks of uprooting by wave action (Schutten, Dainty & Davy, 2005). Exposure of shallow sediments to air can lead to reduction of the organic content of sediments (Degroot & van Wijck, 1993, James et al., 2001) as a result of faster mineralization perhaps due to a higher O2 flux from the atmosphere to support microbial respiration. Higher mineralization could in

turn lead to a larger efflux of liberated nutrients following re-submergence and potentially fuel eutrophication (James et al., 2001, Steinman et al., 2012). Sediments should also become more compact and consolidated following drainage because of loss of organic matter and stronger gravitation leading to better root anchorage upon re-submergence (James et al., 2001). Oxygenation of reduced sediment compounds, primarily sulfide and Fe2+, can, however, lead to acidification following re-submergence (Smolders et al., 2006, Van Wichelen et al., 2007). Despite the potentially downside of increased sediment nutrient release and acidification, it has been proposed to restore sediment conditions by promoting water level fluctuations and periodic drainage of the sediments (Coops & Hosper, 2002). The potential use of water level fluctuations as a restoration tool has, however, not been experimentally investigated in detail for softwater lakes.

We here tested the long-term influence on plant performance and sediment decomposition of permanently aerial, fluctuating aerial-submerged and permanently submerged conditions in controlled laboratory experiments and in field experiments along an elevation gradient with Littorella uniflora inhabiting sandy littoral sediments in Lake Ræv, mid-Jutland, Denmark.

The gradient from aerial to submerged conditions offers changing benefits and constraints to L. uniflora’s performance and sediment’s decomposition rate. At high elevation under drained conditions and air exposure, we predict that L. uniflora should allocate more biomass to roots than leaves to ensure water supply and avoid desiccation. High O2 supply from the atmosphere to surface sediment should also prevent constraints on root development and assist fast decomposition of labile organic matter provided the soil does not dry up entirely. Fast mineralization in turn should release inorganic nutrients and stimulate plant growth in organically enriched sediments more than

85

Isoedtids and water level fluctuations  

in fluctuating or permanently submerged populations. Under water, in contrast, there is obviously no water shortage and associated constraints on leaf development, while possibly reduced oxygenation of sediments may restrict root development and decomposition of organic matter. We therefore predict that proportions of leaves to roots should be high and decomposition should be lower than under drained conditions. Fluctuating exposure to air and water may result in rates of sediment decomposition between those in treatments permanently in air or under water. Fluctuating conditions may represent an extra cost to replace aquatic leaves by aerial leaves upon exposure to air to avoid desiccation, while aerial leaves can be converted to aquatic

leaves upon submergence. We therefore predict that biomass development should be lower under fluctuating conditions than under permanent emergence or submergence. Material and methods

Laboratory experiment

Experimental treatments – The influence of different water level regimes on growth of Littorella uniflora and sediment chemistry and decomposition was tested in cylindrical cores (diameter 10 cm, height 20 cm) over a 160-days long laboratory experiment. Triplicate sediment cores with no plants or with natural densities of L. uniflora were subjected to either submerged, drained or fluctuating water level (Fig 1A).

Fig. 1. Schematic drawing of laboratory setup. A) The six different treatments: planted and bare sediments exposed to either submerged (1), fluctuating (2) or drained (3) conditions. Fluctuating treatments alternated between submerged and drained conditions. B) DIC flux measurements in submerged (4) and drained (5) treatments. Under submerged conditions the flux was measured as accumulation of DIC within the chamber connected to the sediment core, by retrieving small water-samples with a syringe and analysing them in an IRGA system. Under drained conditions the CO2-flux was measured directly in the air of a larger chamber covering the whole sediment core and connected to a portable IRGA.

86

Isoedtids and water level fluctuations  

Plants and sediment were collected in shallow water (0.5 m) in the oligotrophic Lake Ræv (56.02521 lat., 9.443243 long.) in July 2011. Plants were brought back to the laboratory undisturbed in natural sediment turfs (25 cm diameter and 20 cm deep). Turfs were transferred from the cylinder to buckets and kept submerged in lake water until used for experiments within 8 days. Surface sediment (0-20 cm) was also brought to the laboratory where stones and large pieces of organic matter were removed by sieving through a 0.5 cm mesh. Thereafter, a small amount of standard dry pellets of organic pasture grass was crushed, added to the sediment and carefully homogenized. Pellets contained (as a percentage of DW) 91 ± 1% (n = 4) organic matter, 46.9 ± 0.4% organic carbon (C), 2.3 ±0.15% total nitrogen (TN) and 0.27 ± 0.01% total phosphorus. Added organic matter is less nutrient-rich than plankton particles but will stimulate sediment metabolism (Møller & Sand-Jensen, 2011) and elevated the sediment organic content from 0.38 % to 0.54 % of the dry weight. Sediment was then left waterlogged for one week prior to experiments to allow new sediment chemistry to establish. Each experimental core was then filled with 3 kg of homogenized sediment. In half of the cores twenty Littorella uniflora (2550 pants m-2) were planted in equally distributed pre-made holes in the sediment (6 cm deep and 0.5 cm wide). Sediment cores were then gently vibrated to allow sediment consolidation around the roots. Cores with no plants were subjected to similar treatment to avoid initial differences in sediment density. The cores were placed in separate aquaria (9 litre) with water levels being either high (10 cm above the sediment surface) in the flooded treatments, low (15 cm below the sediment surface) in the drained treatment or alternating between these two situations in the fluctuating treatment. The set-up was placed in a temperature controlled room at 14oC in an 12 hour light; 12 hour dark cycle with an incident photosynthetically active irradiance (PAR) of 169 µmol photons m-2 s-1 when

submerged and 184 µmol photons m-2 s-1 when drained. Drained conditions were secured by drilling two 5 mm holes 2 cm above the bottom of these cores before filling it with sediment. Due to capillary forces of sediment particles inside the cores water was drawn in from the aquarium yielding moist but un-flooded conditions at the sediment surface. Drainage holes were covered on the inside with a fine-meshed plastic fabric to allow water to pass but holding sediment particles within the core. When cores were handled during the experiment inserted rubber stoppers prevented leakage.

Because of the large volume of water needed for the experiment we used filtered lake water from nearby Lake Esrum diluted 20 times with demineralized water to obtain chemical condition similar to those of the softwater, oligotrophic Lake Ræv. The incubation medium contained: 0.2 mM dissolved inorganic carbon (DIC), 6.8 µM total nitrogen and 0.2 µM total phosphorus. Water level was kept constant in the flooded and drained treatment by replacing evaporated water with demineralized water and replenishing the water column at regular intervals to avoid algal growth interfering with measurements of metabolism and nutrient fluxes. The fluctuating treatment alternated between the flooded and drained situation to simulate a natural community positioned near the water line in a lake with variable water level. The treatment was initiated and completed with a drained period of four weeks duration. In between sediments were flooded for two periods of five weeks duration separated by a drainage period of three weeks. In submerged treatments, water was kept at air saturation by bubbling with atmospheric air. Plant morphology and performance Plant morphology and leaf chlorophyll were measured before and after the laboratory experiment. Initially, twenty individuals were sacrificed and analysed and after the experiment all plants were counted, analysed

87

Isoedtids and water level fluctuations  

morphologically and leaves were used for measurements of chlorophyll and photosynthesis (only measured at the end of the experiment).

Plants of each replicate core were separated into leaves, stem and roots and dried (105 oC) and weighed and a subsequently combusted (550oC) and biomass calculated as organic dry weight. Values are presented as organic dry weight of leaves and roots (roots and stems).

A subsample of dried leaves were homogenized and analysed for the content of organic carbon (C) on a CHN EA1108-elemental analyser (Carlo Erba Instruments, Milan, Italy). This enabled calculation of the net carbon incorporation of the plants during the experiment using the same proportion of carbon to organic dry weight for the roots as for leaves after correction for mineral content.

Maximum photosynthesis of leaves were measured after the experiment according to the method of Møller and Sand-Jensen (2011) with small modifications. In short, leaves were split lengthwise (to secure efficient exchange of CO2 between leaf and water) and incubated in water in closed glass flasks (volume 50 ml) containing 1000 µM free CO2 (pH 6.9) to ensure inorganic carbon saturation. The flasks were mounted on a rotating wheel and exposed to 340 µmol photons m-2 s -1 for 2 hours at 15 oC. Oxygen concentration inside the flasks was measured with a Firebox 3 O2 optode (PreSens, Regensburg, Germany) after incubation and compared to similarly treated blanks with no leaves. Photosynthetic rate was normalized to leaf dry weight measured after freeze drying.

Chlorophyll a was measured by extraction in ethanol for 24 h and subsequent spectrophotometric analysis (Christoffersen & Jespersen, 1986) of subsamples of the leaves used for photosynthetic measurements. Carbon efflux

Release of organic carbon from the cores was assessed by analysing water samples from each water column taken before and after

renewal of water. Dissolved and particulate organic carbon were measured according to Kragh and Søndergaard (2004).

Efflux or uptake of inorganic carbon from cores was measured midway into the light and dark period several times during the experiment. Under drained conditions, CO2 evolution was measured by inclosing the turf in a large transparent polyethylene cylinder (3.68 litre) preventing contact to the atmosphere (Fig. 1B). The cylinder was connected to a portable IRGA (LI-820, LI-COR biosciences, Lincoln, USA) and a CTS diaphragm pump (Hargraves Technology Coorporation, Mooresville, USA) delivering 1 litre of air per minute through gas impermeable tygon-tube (CM scientific, Silsden, UK). The IRGA was connected to a laptop computer for data logging. Concentration change of CO2 over time and knowledge of volume of chamber and tubing enabled calculation of the CO2 exchange rate.

In the submerged treatments, DIC exchange was determined by mounting a transparent, water-filled chamber (0.785 litre) on top of the sediment core. The chamber had a flat top equipped with a small hole for extracting minute water samples for DIC analysis (Fig. 1). The hole was closed by a rubber stopper between sampling. The inside of the chamber was equipped with a 1 cm stirrer bar rotated by a large magnet outside the chamber to ensure water mixing. DIC exchange was determined as the concentration change during 20-30 minutes and measured on triplicate 100 µl water samples injected into 3 % HNO3 in a bubble chamber purged with N2-gas carrying evolved CO2 into an Infrared gas analyser (IRGA, ADC-225-MK3, Hoddesdon, UK) (Vermaat & Sand-Jensen, 1987).

All measurements of inorganic C flux were performed in both light and darkness three times during the 160-days experiment enabling calculation of the diurnal mean net CO2 evolution. The flux from the fluctuating treatment was measured four times to obtain values for both submerged periods. All rates

88

Isoedtids and water level fluctuations  

were expressed relative to sediment surface area. Sediment geochemistry - After the experiment, a 15 cm long sediment core was retrieved with a Perspex tube (5 cm diameter) from each core. Cores were sliced in depth intervals of 0-2 cm, 2-6 cm and 6-10 cm. Sediment water content was measured as weight loss after 48 hours at 105oC. The dried and homogenized sediment was used for determination of organic matter content as loss on ignition at 550oC. Total iron was measured on combusted samples using the phenanthrolin method modified by Møller and Sand-Jensen (2008). To include the potential sink of iron precipitating on plant roots, a subsample of roots from each vegetated core was combusted at 550oC and iron in the residue was dissolved in HCl and measured as described above.

The amount of reduced iron (Fe2+) adsorbed to sediment particles and dissolved in the pore-water was determined on samples taken with a small perpex tube (1 cm in diameter) split into the same depth intervals as described above. The samples were transferred to 300 ml gas tight glass bottles containing an anoxic 1 M MgCl solution and left for one hour on a shaking table for adsorbed Fe2+ to be fully exchanged with Mg++. Subsequently, samples were left for particles to settle for one hour and the concentration of dissolved Fe2+ in the supernatant was measured as described above. Field experiment

To evaluate how naturally drained and flooded areas with dense vegetation of L. uniflora responded to water level fluctuations and addition of organic matter, a 100-days long field experiment was set up in late summer to autumn (18th July to 25th October 2011) along a steep hydrological gradient in Lake Ræv. The water level in Lake Ræv has in recent years undergone a systematic drawdown of 0.7 m but was relatively constant during the experiment (see below).

Six plots (36*60 cm) were distributed perpendicular to the coastline with the deepest plot located 36 cm below the water and the highest plot elevated 71 cm above the water line at the beginning of the experiment. Each plot was divided into six subplots (12*12 cm) half of which were randomly assigned to organic enrichment while the other half served as controls. Organic matter (1000 g DW m-2) was added as feeding pellets inserted 4 cm into the sediment using a forceps, while in control subplots the forceps was inserted in a similar way to control for the confounding effect of physical disturbance. The subplots were separated by buffer zones ensuring that leakage from organic subplots did not affect the controls. The addition of organic matter was equivalent to 0.69±0.05 % extra organic matter relative to sediment dry weight in the upper 0 -10 cm of the sediment. According to regular water level recordings the driest plot had been emerged for 2 years, the second driest plot for 1.5 years and the third driest plot for one year, whereas the fourth plot had been subjected to fluctuating water for one year prior to the experiment.

0.0 0.2 0.4 0.6 0.80

10

20

30

40

50

Elevation (m)

Sed

imen

t moi

stur

e (%

)

Fig. 2. Soil moisture (SM, %, mean ± SD) in

the upper 6 cm of the terrestrial plots in the field along the elevation gradient (E, m). Each group of points represent measurements in a given plot at three occasions during the investigation period. Small changes in water level between measurements resulted in small changes in position of plots along the gradient. The linear regression line (SM =-61.6 x E + 51.1) was highly significant (p<0.001, r2= 0.85).

89

Isoedtids and water level fluctuations  

The water level was measured by a pressure sensitive water level logger (HOBO U20) placed in the lake and a reference logger placed above the water. The water level gradually rose by 10 cm during the 100-days long experiment flooding the lowest terrestrial plot 80 days into the experiment. The terrestrial plots covered the entire zone in which L. uniflora grows terrestrially whereas it extends to a water depth of about 2 m. At regular intervals throughout the experiment, sediment moisture was measured in four replicates within each terrestrial plot with a Theta Probe ML2x (Delta-T Devices Ltd, Cambridge, UK). The probe was inserted into the buffer zones between subplots to prevent plant damage. The sediment moisture of the plots did not change markedly during the experiment until the final sampling date where the lowest terrestrial plot had become submerged (Fig. 2). Sediment moisture content (SM, %) was closely related to elevation (E, m) of each plot relative to the water line (linear regression p< 0,001 r2=0.85, SM =-61.6 E + 51.1). To allow inclusion of the two submerged plots and because of low variation in soil moisture during the experiment, the entire flooding gradient was expressed as average elevation relative to the water line; positive values indicated aerial conditions.

At the end of the experiment, a sediment sample (5 cm in diameter, >10 cm long) was retrieved from each subplot in Perspex tubes. Each tube were closed with rubber stoppers to avoid disturbance and brought to the laboratory for analysis. The rest of each subplot was gently excavated collecting all plant material. Plants were placed in sealed plastic bags with moist paper towel, kept cool and brought to the laboratory.

In the laboratory, sediment samples were split and analysed for content of water, organic matter and carbon using the same procedure as for sediments from the laboratory experiment. Knowing the organic matter content of control and enriched plots enabled estimation the sediment carbon loss

along the elevation gradient during the experiment period. The loss was calculated as the difference between enriched and control plots subtracted from the amount initially added to sediments. The sediment layer from 2 to 10 cm depth was used in the calculation to avoid the effect of mosses and algae often growing on sediment surfaces and it was assumed that carbon constituted the same proportion of organic matter in initial sediments as in the added organic matter (46.9 %).

All plants from each subplot were counted, split into leaves and below-ground biomass (roots and stem) and the dry weight recorded. The organic dry weight was determine by correcting for the mineral content of submerged plants collected at the site in July showing a low content of both roots and shoots.

Leaves collected at the end of the experiment were analysed for organic C content using the same method as for leaves in the laboratory experiment. As the mineral content of leaves and roots was similar organic C was assumed to have the same proportion of dry weight in above and below ground biomass. Calculating the difference in plant C between control and enriched plots enabled estimation of the proportion of C from sediment mineralization of added organic matter being incorporated in new plant biomass.

Photosynthesis was determined for leaves of the collected plants using the method described for laboratory experiments. Statistical analysis

Statistical analysis and graphs were performed in Graph Pad Prism 5 (Graph Pad Software Inc., La Jolla, USA). Differences between single variables in treatments in the laboratory experiment and along the elevation gradient in the field were tested with one-way ANOVA followed by Tukeys test. The effect of the treatments on two variables (e.g. root and shoot biomasses) were analysed by two-way ANOVA on joined data from treatments

90

Isoedtids and water level fluctuations  

0

100

200

SubFluc

Drain

start

Laboratory

Water regime

Bio

mas

s (g

DW

m-2

)

-0.5 0.0 0.5 1.0

Field

Elevation (m)

Fig. 3. Above- and below-ground biomass (white and black bars, mean ± SD) relative to surface area of Littorella uniflora in the laboratory experiment and field investigation (control plots only). In the laboratory experiment the total plant biomass increased in all treatments from the start values. Under drained conditions however, plants grew significantly more compared to submerged and fluctuating conditions (one-way ANOVA, p<0.05). In the field experiment, the total biomass of the plot immediately below and the two plots above the water line was significantly larger than the two direst plots (one-way ANOVA, p<0.05). For detailed interpretation on biomass allocation to above- and below-ground biomass see results. with and without plants followed by Bonferroni post-tests. Fmax-tests were used to evaluate variance homogeneity before parametric tests. In a few instances data had to be log transformed to meet the requirements for variance homogeneity. The significance of differences between initial plant biomass of the cores and biomass after the experiment was evaluated by assessing the overlap of 95 % confidence limits, because of the very low variance of initial biomass. Values are presented as ± standard deviation (SD). Results

Plant response to water level

Laboratory experiments - Plant biomass increased during the 160-days long experiment in all treatments, but it became much higher under drained than under submerged or fluctuating conditions (Fig. 3, one-way ANOVA, p<0.001). This high biomass under drained conditions was primarily caused by a significantly larger root

biomass under drained conditions, while leaf biomass did not differ between treatments (Fig. 3, two-way ANOVA, p<0.001). Because plant density was lowest in the drained treatment, the high biomass was caused by individual plants with significantly larger leaf mass and, in particular, high root mass compared to plants from submerged and fluctuating conditions (Fig. 4, two-ANOVA, p<0.001). Plants biomass in the fluctuating treatment tended to be lower than in the submerged treatment and a t-test of only these two treatments revealed a significant difference (t-test, p=0.033). This finding showed that fluctuating conditions had a negative influence on growth of L. uniflora though the biomass still increased during the experiment. At the end of the experiments, plants in the fluctuating water regime had been in air for 28 days after the last submergence and plants held both aerial and submerged leaf types (sensu Hostrup & Wiegleb, 1991). The two leaf types had leaf chlorophyll contents very similar to those of leaves from permanently submerged and drained

91

Isoedtids and water level fluctuations  

-0.5 0.

00.5

1.0

Field

Elevation (m)

0.00

0.05

0.10

Laboratory

SubFlu

cDra

in

Water regime

Indi

vidu

al b

iom

ass

(g D

W p

lant

-1)

Fig. 4. Above- and below-ground individual plant mass (white and black bars, mean ± SD) from the different

water regimes in the laboratory experiment and in the field experiment (control plots) along the elevation gradient. In laboratory experiment the whole plant biomass was significantly higher under drained conditions (ANOVA, p<0.05), while in the field experiment the two plots immediately above the water had significantly higher biomass than the deepest plot (ANOVA, p<0.05). For detailed interpretation on biomass allocation to above- and below-ground biomass see results. conditions and did not differ significantly from those (Fig. 5, one-way ANOVA p<0.001, Tukey p>0.05). For photosynthesis, the aerial leaves from the fluctuating conditions had significantly higher rates than all other treatments (Fig. 5, one-way ANOVA p<0.001, Tukey p<0.05) and both photosynthetic rate and leaf chorophyll content suggested a limited direct stress on the plants due to the fluctuations. Field experiments -The individual plant weight did not differ between emergent plots along the natural gradient in elevation except the second highest plot and soil moisture having a lower plant mass than the plot located just above the water line (one-way ANOVA, p<0.05). Plants in the remaining terrestrial plots were all significantly heavier than in the deepest submerged plots (Fig. 4, Tukey, p< 0.05). Biomass allocation to both leaves and roots differed significantly along the gradient (two-way ANOVA, p<0.05). A strong significant interaction indicated that increasing elevation resulted in opposing trends in biomass allocation, while root

biomass increased to shoot biomass tended to decrease (Fig. 4, p<0.05). Leaf photosynthesis under standard conditions did not differ significantly along the elevation gradient (Fig. 5, one-way ANOVA, p=0.57), even though chlorophyll content was significantly higher in the submerged treatment (Fig. 5, one-way ANOVA, p< 0.001, Tukey, p<0.05).

The high individual plant weight in most dry plots did not lead to a higher biomass per surface area because of low plant density (Fig. 3). Instead, the highest biomass was found at intermediate plant density close to the water line, and the two driest plots had a significantly lower biomass compared to all other plot except the deepest plot from which they did not differ (one-way ANOVA, p<0.05, Tukey, p<0.05). The field response generally showed high resemblance to the response in the laboratory experiment, emphasizing the validity of the latter. Plant response to organic enrichment in

the field experiment

Field experiments – Influence of added organic matter on plant performance relative to control plants in natural sediments was

92

Isoedtids and water level fluctuations  

Start

WL

Sub W

L

Fluc W

L

Fluc A

L

Drain A

L0

5

10

Laboratorya aa

b b

*

Water regime

Le

af ch

loro

ph

yll (

mg

g-1

DW

)

-0.5 0.0 0.5 1.0

0

500

1000Field

a

bb bb

Elevation (m)

NP

(µmol O

2 g-1 D

W h

-1)

Fig. 5. Leaf chlorophyll (black bars) and net photosynthetic rate (white bars) of leaves in laboratory and field experiments (control plots) along the elevation gradient (mean ± SD). Plants from the fluctuating treatments carried both water leaves (WL) and aerial leaf types (AL). Significant differences in chlorophyll content are indicated by letters while differences between photosynthetic rates are indicated by asterisk. The two leaf types of the fluctuating plants highly resembled those of their counterparts in the two other treatments. only tested in the field experiment. Addition of labile organic matter increased biomass across treatments (Fig. 6A, two-way ANOVA, p<0.05). Along the elevation gradient the enrichment effect was significantly stronger for leaves than for root development (linear regression, p=0.03). The addition of organic matter did not affect the water content of the sediments, while there was a strong effect of elevation (two-way ANOVA). Thus, the increased plant growth in organically enriched sediments was more likely due to higher sediment concentration of CO2 and nutrients due to higher decomposition rates. Sediment conditions and carbon budget

Laboratory experiment - Different water regimes and plant cover in the laboratory experiment strongly affected carbon metabolism and, thus, the flux of inorganic and organic carbon from the sediment to the surrounding medium during the 160 days. All treatments lost more sediment carbon than they gained by plant uptake of DIC from the

-0.5 0.0 0.5 1.0

1

2

3

4leavesroots

B

Elevation (m)

Rel

ativ

e bi

omas

s in

crea

se

0

100

200

300

400A

Bio

mas

s (g

DW

m-2

)

Fig 6. Field experiment. Effect of addition of sediment organic matter on plant biomass (mean ± SD) across the investigated elevation gradient. A) Addition of organic matter significantly increased the biomass (black bars) compared to control plots (white bars) across the entire gradient (two-way ANOVA, p<0.05). B) The relative increase in biomass increased markedly under aerial conditions and more for leaves than roots.

93

Isoedtids and water level fluctuations  

Table 1. Laboratory experiment. Sediment water content, loss of sediment organic carbon, efflux of DIC and TOC from the sediment, and increase of plant biomass (mean values ± SD) during the 160-days experiment. Differences were tested by one-way ANOVA followed by Tukeys-test and significant differences are indicated by different letters. Treatment Plants Sediment water

content (% of WW)

DIC flux (mol m-2)

Plant growth (mol m-2)

DIC +plant growth (mol m-2)

TOC flux (mol m-2)

Submerged + *19.6±0.5 4.4±0.6a 3.5±0.06a 7.8±0.6ab 4.3±0.7a

Submerged - *18.9±0.2 3.1±0.9a 3.1±0.9a 3.2±0.2ab

Fluctuating + 15.1±2.4 4.8±0.8a 2.4±0.5a 7.2±0.9a 2.3±0.2b

Fluctuating - 15.0±2.6 11.4±2.2bc 11.4±2.2bc 2.2±0.5b

Drained + 12.7±2.6 7.9±3.3ac 7.8±1.0c 15.7±2.8cd

Drained - 13.4±4.6 17.1±2.3d 17.1±2.3d

*flooded sediments water as indicated by a significant reduction of organic matter in all treatments compared to initial sediment conditions (one-way ANOVA, p<0.05, Tukey p<0.05). Bare sediments showed a profound and highly significant increase of the DIC efflux from submerged (3.1±0.9 mol C m-2), over fluctuating (11.4±2.2 mol C m-2) to permanently drained conditions (17.1±2.3 mol C m-2, Table 1, one-way ANOVA, p<0.05, Tukey p<0.05). In vegetated sediments part of the CO2 released by mineralization in the sediment can be incorporated into the growing plant biomass which can account for the reduced DIC efflux observed under fluctuating and permanently drained conditions compared to bare sediments. The sum of DIC efflux and accumulation of plant biomass, however, was at the same level in bare and vegetated sediments under fluctuating (11.4±2.2 and 7.2±0.9 mol C m-2, respectively) and permanently drained conditions (17.1±2.3 and 15.7±2.8 mol C m-2, respectively). These combined rates can be regarded as suitable measures of total sediment mineralization to CO2 and they were much lower under permanently submerged conditions with either bare (3.1±0.9 mol C m-

2) or plant covered sediments (7.8±0.6 mol C m-2). Under submerged and fluctuating conditions dissolved organic carbon (DOC)

can be lost from sediments and dead leaves to the water. This loss of organic carbon was particularly high in the submerged treatment with plants, and lower in the submerged treatments without plants. Correcting for the shorter period of submergence in fluctuating treatments, these treatments had a higher daily efflux of organic matter during submergence supporting the hypothesis that a burst of organic material degraded to low molecular compounds under drained conditions is released when submergence follows emergence. When the DOC loss was included in the efflux of DIC and the increase of plant biomass, the total loss of carbon from the sediments still remained lower in submerged vegetated (13.1 mol C m-2) and bare sediments (6.3) than in permanently drained sediments (17.7 and 17.0).

The larger efflux of DIC and the larger sum of DIC efflux and increase of plant biomass under drained and fluctuating treatments compared with flooded treatments support the hypothesis that sediment mineralization is higher under the former conditions. The low proportion of reduced Fe2+ in sediments of drained and fluctuating treatments is in accordance with the prevalence of oxygenated conditions supporting high and complete mineralization to CO2 under these conditions (Fig. 7). The

94

Isoedtids and water level fluctuations  

Table 2. Field experiment. Water content, loss of organic carbon content and increase in plant biomass (mean values ± SD) in organically enriched sediments relative to control sediments along the elevation gradient during the 100-days experiment. Differences were tested by one-way ANOVA followed by Tukeys-test and significant differences are indicated by different letters. Plot elevation (m)

Sediment water content (% of WW)

Org carbon loss (mol C m-2)

Plant biomass increase (mol C m-2)

-0.39 24.5*±1.8 27.5±4.6ab 0.8±1.1a

-0.18 30.6*±5.2 29.4±5.6ab 1.8±2.5ab

0.05 31.2*±1.4 23.0±1.3ab 1.3±1.8a

0.27 22.9±1.7 22.7±3.4ab 6.0±0.5b

0.46 9.4±0.4 32.6±0.9a 1.5±0.3a

0.68 5.4±0.4 36.3±1.5a 3.4±1.9ab

*flooded at sampling effect of both water regime and plant cover was significant, with the former explaining the majority of the variation (two-way ANOVA, p<0.05).

Sediment density (g DW cm-3) was significantly higher in the top layer (0-2 cm) in drained and fluctuating sediments compared to submerged sediments in treatments with no plants (Fig. 8, two-way ANOVA, p<0.05). Sediment density increased significantly with depth in the submerged treatment, whereas density was constant and high in fluctuating and drained sediments (two-way ANOVA, p<0.05). The same trends between treatments and with depth were observed in sediments with plants but different sediment volume occupied by roots did not allow accurate quantification. Field experiments – In the field a large fraction of the 39 mol organic C m-2 added to the sediments was lost during the 100-days experiment (Table 2). The loss ranged from 22.7 to 36.3 mol C m-2 and it was highest in the driest plots followed by the submerged plots while the smallest quantity was lost immediately above the waterline. The difference between plant biomass in control and organically enriched plots along the elevation gradient accounted for part of the

Submerged Fluctuating Drained0

5

10

15

% r

educ

ed F

e

Fig. 7. Laboratory experiment. The percentage of reduced Fe2+ of the total Fe pool in the sediment (mean ± SD) subjected to the different water regime and plant cover (white bars) or no plant cover (black bars). Lower percentage of reduced Fe2+ indicates more oxygen available in sediments. A two ANOVA showed significant effect of both water regime and plant cover (p<0.05). CO2 released from organic matter. This amount was highest (6 mol C m-2) in the middle of the investigated gradient. Compared to the laboratory experiment the proportion of the degraded organic C being incorporated into plants was substantially lower in the field. This difference could result from enriching the sediment at a single depth rather than exposing plant roots to extra organic matter distributed homogeneously over all depth as in the laboratory experiment. Discussion

Plant response to water level regimes

Measurements confirmed that Littorella uniflora is well adapted to amphibious life at the margins of oligotrophic, softwater lakes. Plant biomass increased under drained, fluctuating as well as submerged conditions in laboratory experiments and along the elevation gradient from water to air in the field and leaf chlorophyll and photosynthesis remained high under all treatments. Other amphibious species, which are dependent on CO2 uptake through leaf surfaces for photosynthesis, can survive under water but usually have very low photosynthesis and lose weight when the water contains low CO2

95

Isoedtids and water level fluctuations  

concentrations close to air saturation (Sand-Jensen et al., 1992). Mentha aquatic, for example, has a negative relative growth rate (-0.8 d-1) in air-saturated water and it requires 15-fold super-saturation of CO2, which can be found in groundwater-fed lowland streams, to support appreciable growth rates (1.8 d-1), though growth rates are three times higher in air (4.7 d-1, Sand-Jensen & Lindegaard, 2004, their Table 6.2 ). Littorella uniflora and other isoetid species grow well in oligotrophic, softwater lakes with low CO2 concentrations close to air saturation in the water because they exploit the CO2-rich sediment resources (Raven et al., 1988, Pedersen, Sandjensen & Revsbech, 1995). The small isoetid species can maintain dominance under fluctuating water levels at the margins of oligotrophic, softwater lakes because of low competition from taller species which require higher nutrient levels, are less tolerant to the flooding regime and wave exposure, are unable to utilize the CO2-rich hydrosoil and have no access to alternative inorganic carbon sources (e.g. HCO3

-) under water (Farmer & Spence, 1986). There was no indication in our experiments that enrichment of sediments with labile organic matter stressed Littorella uniflora. On the contrary, plant growth was greatly stimulated in the field presumably due to higher sediment mineralization supplying extra CO2 and inorganic nutrients for plant uptake (Sand-Jensen & Søndergaard, 1979, Bagger & Madsen, 2004). Previous laboratory experiments with submerged L. uniflora have documented that enrichment of sediments with organic matter of the same type and amount (0.1 and 0.2 % of sediment DW) stimulates plant growth, while greater organic enrichment (0.4-1.6 %) induces sediment anoxia, reduces plant growth, chlorophyll and photosynthesis and leads to stunted roots having difficulties receiving sufficient O2 flux to the apical root meristems and supplying the leaves with adequate nutrients (Raun, Borum & Sand-Jensen, 2010, Møller & Sand-Jensen, 2012).

It is likely, that L. uniflora under drained or alternating drained and submerged conditions can better sustain high organic enrichment because exposure to air will prevent leaf anoxia, assist in sediment oxygenation and also lead to lower proportions of reduced Fe2++, and other reduced potential phytotoxins (e.g., Mn2+, NH4+ and small fatty acids; Armstrong & Armstrong, 2001, Smolders et al., 2002) characteristic of anoxic sediment conditions (Fig. 7). The highest biomass in laboratory experiment was found under drained conditions and, likewise, the largest plants grew in the dry part of the elevation gradient in the field. This finding is in accordance with previous field investigations of L. uniflora showing higher growth rates above than under water (Robe & Griffiths, 1998). Although individual plants in the field were larger under drained conditions, the biomass was smaller in the driest plots than in the plots located close to the water line having higher plant density.

1.4 1.5 1.6 1.7 1.8

0

2

4

6

8

10

Sediment density (g DW cm-3)

De

pth

(cm

)

Fig. 8. Laboratory experiment. The sediment density of un-vegetated sediments (mean ± SD) of the three different water level regimes. Black squares: submerged, white circles: fluctuating and black circles: drained conditions after the laboratory experiment. The sediments were significantly less condensed in the top layer under submerged conditions compared to the other two conditions. Also the sediment density increased with depth (two-way ANOVA, p<0.05).

96

Isoedtids and water level fluctuations  

The lower plant density at the highest elevation, which had been out of the water for more than a year prior to the investigation, can be explained by poor germination because lack of moist sediments and prolonged seed desiccation impair germination (Arts & van der Heijden, 1990) and by lack of vegetative spread by runners under dry conditions (Robe & Griffiths, 1998).

The fluctuating water level in laboratory experiments was accompanied by reduced biomass development relative to permanently emerged or submerged conditions. This finding suggests that fluctuations were costly and reduced biomass development although leaf chlorophyll and photosynthesis were not depressed and did not suggest physiological stress (Møller & Sand-Jensen, 2011). Plants under fluctuating conditions had both aerial and aquatic leaf types conforming to earlier results showing that emerged aquatic leaves can remain active for more than a month (Robe & Griffiths, 1998). While the aquatic leaves can remain active under emerged conditions they are certainly not as suitable as the terrestrial leaves as they lack stomata for CO2 uptake and have a thin cuticle resulting in a high water loss. The lower biomass under fluctuating conditions is probably due the extra costs linked with higher leaf turnover. Aquatic submerged leaves of L. uniflora remain active for six-nine months before they are shed (Sand-Jensen & Søndergaard, 1978), while in the present experiments plants are inclined to produce new leaves every five weeks because of the shifting water level.

Leaf and root development of L. uniflora responded in different ways to drainage and flooding. Root mass increased markedly relative to leaf mass under drained conditions in both laboratory and field experiments probably in response to limited water and nutrient availability. This response of increasing biomass allocation to the organ exploiting the most limited resources is widespread among plants (Poorter & Nagel, 2000) and lower root to shoot biomass is also

typical of submerged aquatic plants compared with terrestrial plants (Wetzel, 2001). The same root-shoot response as we found for L. uniflora has been observed in earlier Danish work (Nielsen & Sand-Jensen, 1997), while the root-shoot ratio remained constant upon emergence in Scottish studies (Robe & Griffiths, 1998). This difference could be caused by the Scottish plants being regularly exposed to air and therefore, already possessing long roots because of earlier periods of emergence. In the Danish study the submerged plants had not been exposed to air for several years and lacked the special long (15-20 cm), robust roots that can absorb water from deep sediments.

Sediment response to water level regimes

While all sediments exposed to different water level regimes in the laboratory experiment lost organic matter, the amount differed significantly between treatments. The largest direct DIC efflux was observed in the drained treatment without vegetation but if incorporation of carbon into plant biomass was accounted for the combined loss from the drained vegetated sediments was of similar magnitude. This finding corresponds with earlier investigations showing high loss of organic matter under drained conditions (Degroot & van Wijck, 1993, van Wijck & Degroot, 1993, James et al., 2001). The lowest loss of organic matter was observed in submerged bare sediments, while the submerged vegetated and both treatments with fluctuating water levels lost intermediate amounts. The low loss of the submerged bare sediment is likely caused by lower mineralization rates due to reduced oxygen depth penetration in organically enriched sediments (Møller & Sand-Jensen, 2011) as also supported by the high proportion of reduced Fe2+. It is possible that there was a higher build-up of DIC in the sediment in this treatment because there were no plant roots to incorporate CO2 in plant biomass or serve as an air channel conduit for transport from the sediment to the water. However, the

97

Isoedtids and water level fluctuations  

experiment lasted so long (i.e.160 days) that this sediment accumulation must be low relative to the measured efflux. However, it is possible that part of the anoxic respiration instead of forming CO2, released methane which we did not quantify. If so, the carbon efflux would have been underestimated. In the field experiment, the decline of added organic matter was measured directly by weight differences between enriched and control sediments. This decline was also highest under the driest conditions, but lowest in intermediate elevations, while submerged plots were located in between. In the field, we did not measure the DIC efflux. The substantial loss in the submerged plots could be due to release of organic matter from the freshly added dry pellets that are dissolved under these conditions, while this process will not take place to nearly the same extent in naturally formed aquatic detritus. Restoration of isoetid populations

Seepage lakes is a preferred habitat of isoetid species because they are nutrient-poor and experience widely fluctuating water levels that are tolerated and perhaps even preferred by isoetid species (Sculthorpe, 1967), while competitors among aquatic elodeid species dry out upon emergence and taller amphibious or terrestrial species are unable to photosynthesize and grow under water (Sand-Jensen et al., 1992, Sand-Jensen & Frost-Christensen, 1999, Bailey-Serres & Voesenek, 2008). It is, therefore, a problem when the water level of oligotrophic lakes becomes strictly regulated because isoetid species then lose their competitive advantage of being able to photosynthesize and grow well in the fluctuating transition zone between air and water (Pedersen & Sand-Jensen, 1992, Nielsen & Sand-Jensen, 1997).

There are other advantages for isoetids linked to the influence on sediment properties of fluctuating water levels. This is particularly the situation when high sedimentation of organic matter due to external input or high in-lake phytoplankton production threatens to

make the sediments unsuitable to growth and root anchorage of isoetids by preventing oxygenation, increasing formation of potential plant toxins and making the texture less firm and cohesive (Møller & Sand-Jensen, 2008, 2011, 2012). Then, exposure to air can serve, as documented here, to increase the oxygen flux, degradation of organic matter, plant growth and specific density of surface sediments. This did not lead to problems with acidification (Smolders et al., 2006) in our study (data not shown) because FeS levels were low in the sediments. Lower water content and higher specific density, in turn, are known to improve root anchorage (Schutten et al., 2005, Sand-Jensen & Møller, 2013).

In those instances where thick surface layers of muddy organic matter prevent isoetid establishment and growth, a promising, though costly, restoration tool is to combine reduced water level with direct removal of the drained organic surface layer once again exposing the underlying solid mineral sediment to isoetid reestablishment. This procedure has been used with success in several lakes in the Netherlands (Brouwer, Bobbink & Roelofs, 2002) and with the same effect but less well documented in some Danish lakes (Miljøcenter-Ribe, 2008).

As sediment desiccation has also been shown to decrease the level of nitrogen in sediments (James, Barko & Eakin, 2004, Smolders et al., 2006), this tool may prove valuable in lakes where sediments accumulation has not been too excessive.

Acknowledgements

This project was funded by the Centre for Lake Restoration (CLEAR) a Villum Kann Rasmussen Centre of Excellence.  

References

Armstrong, J. & Armstrong, W. (2001) Rice and Phragmites: effects of organic acids on growth, root permeability, and radial oxygen loss to the

98

Isoedtids and water level fluctuations  

rhizosphere. American Journal of Botany, 88, 1359-1370.

Arts, G.H.P. & Denhartog, C. (1990) Phytogeographical aspects of the West European soft-water macrophyte flora. Acta Botanica Neerlandica, 39, 369-378.

Arts, G.H.P. & Van Der Heijden, R.a.J.M. (1990) Germination ecology of Littorella uniflora (L) Aschers. Aquatic Botany, 37, 139-151.

Baastrup-Spohr, L., Sand-Jensen, K., Nicolajsen, S.V. & Bruun, H.H. (2013) From pond to alvar: predicting plant community composition along a steep hydrological gradient. Journal of Ecology, In revision.

Bagger, J. & Madsen, T.V. (2004) Morphological acclimation of aquatic Littorella uniflora to sediment CO2 concentration and wave exposure. Functional Ecology, 18, 946-951.

Bailey-Serres, J. & Voesenek, L. (2008) Flooding stress: Acclimations and genetic diversity. Annual Review of Plant Biology, 59, 313-339.

Baldwin, D.S. & Mitchell, A.M. (2000) The effects of drying and re-flooding on the sediment and soil nutrient dynamics of lowland river-floodplain systems: A synthesis. Regulated Rivers-Research & Management, 16, 457-467.

Brock, M.A., Nielsen, D.L., Shiel, R.J., Green, J.D. & Langley, J.D. (2003) Drought and aquatic community resilience: the role of eggs and seeds in sediments of temporary wetlands. Freshwater Biology, 48, 1207-1218.

Brouwer, E., Bobbink, R. & Roelofs, J.G.M. (2002) Restoration of aquatic macrophyte vegetation in acidified and eutrophied softwater lakes: an overview. Aquatic Botany, 73, 405-431.

Casanova, M.T. & Brock, M.A. (2000) How do depth, duration and frequency of flooding influence the establishment of wetland plant communities? Plant Ecology, 147, 237-250.

Christoffersen, K. & Jespersen, A.M. (1986) Gut evacuation rates and ingestion rates of Eudiaptomus graciloides measured by means of the gut flourescence method. Journal of Plankton Research, 8, 973-983.

Colmer, T.D. (2003) Long-distance transport of gases in plants: a perspective on internal aeration and radial oxygen loss from roots. Plant Cell and Environment, 26, 17-36.

Coops, H. & Hosper, S.H. (2002) Water-level management as a tool for the restoration of

shallow lakes in the Netherlands. Lake and Reservoir Management, 18, 293-298.

Degroot, C.J. & Van Wijck, C. (1993) The impact of disiccation of a frewater marsh (Garcine Nord, Camargue, France) on sediment water interactions. 1. The sediment chemistry. Hydrobiologia, 252, 83-94.

Engelhardt, K.a.M. (2006) Relating effect and response traits in submersed aquatic macrophytes. Ecological Applications, 16, 1808-1820.

Farmer, A.M. & Spence, D.H.N. (1986) The Growth strategies and distribution of isoedtids in Scottish freshwater lochs. Aquatic Botany, 26, 247-258.

Hostrup, O. & Wiegleb, G. (1991) Anatomy of leaves of submerged and emergent forms of Littorella uniflora (L.) Ascherson. Aquatic Botany, 39, 195-209.

James, W.F., Barko, J.W. & Eakin, H.L. (2004) Impacts of sediment dewatering and rehydration on sediment nitrogen concentration and macrophyte growth. Canadian Journal of Fisheries and Aquatic Sciences, 61, 538-546.

James, W.F., Barko, J.W., Eakin, H.L. & Helsel, D.R. (2001) Changes in sediment characteristics following drawdown of Big Muskego Lake, Wisconsin. Archiv Fur Hydrobiologie, 151, 459-474.

Justin, S.H.F.W. & Armstrong, W. (1987) The anatomical characteristics of roots and plant response to soil flooding. New Phytologist, 106, 465-495.

Keddy, P.A. & Reznicek, A.A. (1986) Great Lakes vegetation dynamics - the role of fluctuting water levels and buried seeds. Journal of Great Lakes Research, 12, 25-36.

Kragh, T. & Søndergaard, M. (2004) Production and bioavailability of autochthonous dissolved organic carbon: effects of mesozooplankton. Aquatic Microbial Ecology, 36, 61-72.

Maberly, S.C. & Madsen, T.V. (2002) Freshwater angiosperm carbon concentrating mechanisms: processes and patterns. Functional Plant Biology, 29, 393-405.

Madsen, T.V. & Sand-Jensen, K. (1991) Photosynthetic carbon assimilation in aquatic macrophytes. Aquatic Botany, 41, 5-40.

Miljøcenter-Ribe (2008) Notat om Vegetationsundersøgelse i Ål Præstesø, NOVANA. Fiskeøkologisk laboratorium, Ribe, DK

99

Isoedtids and water level fluctuations  

Mjelde, M., Hellsten, S. & Ecke, F. (2013) A water level drawdown index for aquatic macrophytes in Nordic lakes. Hydrobiologia, 704, 141-151.

Møller, C.L. & Sand-Jensen, K. (2008) Iron plaques improve the oxygen supply to root meristems of the freshwater plant, Lobelia dortmanna. New Phytologist, 179, 848-856.

Møller, C.L. & Sand-Jensen, K. (2011) High sensitivity of Lobelia dortmanna to sediment oxygen depletion following organic enrichment. New Phytologist, 190, 320-331.

Møller, C.L. & Sand-Jensen, K. (2012) Rapid oxygen exchange across the leaves of Littorella uniflora provides tolerance to sediment anoxia. Freshwater Biology, 57, 1875-1883.

Mommer, L. & Visser, E.J.W. (2005) Underwater photosynthesis in flooded terrestrial plants: A matter of leaf plasticity. Annals of Botany, 96, 581-589.

Nielsen, S.L. (1993) A comparison of aerial and submerged photosynthesis in some Danish amphibious plants. Aquatic Botany, 45, 27-40.

Nielsen, S.L., Gacia, E. & Sand-Jensen, K. (1991) Land plants of amphibious Littorella uniflora (L.) Aschers. maintain utilization of CO2 from the sediment. Oecologia, 88, 258-262.

Nielsen, S.L. & Sand-Jensen, K. (1997) Growth rates and morphological adaptations of aquatic and terrestrial forms of amphibious Littorella uniflora (L) Aschers. Plant Ecology, 129, 135-140.

Pedersen, O., Andersen, T., Ikejima, K., Hossain, M.Z. & Andersen, F.O. (2006) A multidisciplinary approach to understanding the recent and historical occurrence of the freshwater plant, Littorella uniflora. Freshwater Biology, 51, 865-877.

Pedersen, O. & Sand-Jensen, K. (1992) Adaptations of submerged Lobelia dortmanna to aerial life-form - morphology, carbon-sources and oxygen dynamics. Oikos, 65, 89-96.

Pedersen, O. & Sand-Jensen, K. (1997) Transpiration does not control growth and nutrient supply in the amphibious plant Mentha aquatica. Plant Cell and Environment, 20, 117-123.

Pedersen, O., Sandjensen, K. & Revsbech, N.P. (1995) Diel pulses of O2 and CO2 in sandy lake sediments inhabited by Lobelia dortmanna. Ecology, 76, 1536-1545.

Poorter, H. & Nagel, O. (2000) The role of biomass allocation in the growth response of plants to different levels of light, CO2, nutrients and water: a quantitative review. Australian Journal of Plant Physiology, 27, 595-607.

Raun, A.L., Borum, J. & Sand-Jensen, K. (2010) Influence of sediment organic enrichment and water alkalinity on growth of aquatic isoetid and elodeid plants. Freshwater Biology, 55, 1891-1904.

Raven, J.A., Handley, L.L., Macfarlane, J.J., Mcinroy, S., Mckenzie, L., Richards, J.H. & Samuelsson, G. (1988) The role of CO2 uptake by roots and CAM in acquiaition of inorganic C by plants of the isoetid life-form - a review, with new data on Eriocaulon decangulare L. New Phytologist, 108, 125-148.

Riis, T. & Hawes, I. (2002) Relationships between water level fluctuations and vegetation diversity in shallow water of New Zealand lakes. Aquatic Botany, 74, 133-148.

Riis, T. & Sand-Jensen, K. (1998) Development of vegetation and environmental conditions in an oligotrophic Danish lake over 40 years. Freshwater Biology, 40, 123-134.

Robe, W.E. & Griffiths, H. (1998) Adaptations for an amphibious life: changes in leaf morphology, growth rate, carbon and nitrogen investment, and reproduction during adjustment to emersion by the freshwater macrophyte Littorella uniflora. New Phytologist, 140, 9-23.

Sand-Jensen, K. & Frost-Christensen, H. (1999) Plant growth and photosynthesis in the transition zone between land and stream. Aquatic Botany, 63, 23-35.

Sand-Jensen, K. & Lindegaard, C. (2004) Ferskvandsøkologi (in Danish), Gyldendahl, Copenhagen.

Sand-Jensen, K. & Møller, C.L. (2013) Reduced root anchorage of freshwater plants in organic sediments - a lasting risk of eutrophication. In prep.

Sand-Jensen, K., Pedersen, M.F. & Nielsen, S.L. (1992) Photosynthetic use of inorganic carbon among primary and secondary water plants in streams. Freshwater Biology, 27, 283-293.

Sand-Jensen, K. & Søndergaard, M. (1978) Growth and production of isoetids in oligotrophic Lake Kalgaard Denmark. Internationale Vereinigung fuer Theoretusche

100

Isoedtids and water level fluctuations  

und Angewandte Limnogie Verhandlungen, 20, 659-666.

Sand-Jensen, K. & Søndergaard, M. (1979) Distribution and quatitative development of aquatic macrophytes in relation to sediment characteristics in oligptrophic Lake Kalgaard, Denmark. Freshwater Biology, 9, 1-11.

Schutten, J., Dainty, J. & Davy, A.J. (2005) Root anchorage and its significance for submerged plants in shallow lakes. Journal of Ecology, 93, 556-571.

Sculthorpe, C.D. (1967) The biology of aquatic vascular plants, Edward Arnold, London, UK.

Smolders, A.J.P., Lucassen, E.C.H.E.T. & Roelofs, J.G.M. (2002) The isoetid environment: biogeochemistry and threats. Aquatic Botany, 73, 325-350.

Smolders, A.J.P., Moonen, M., Zwaga, K., Lucassen, E., Lamers, L.P.M. & Roelofs, J.G.M. (2006) Changes in pore water chemistry of desiccating freshwater sediments with different sulphur contents. Geoderma, 132, 372-383.

Steinman, A.D., Ogdahl, M.E., Weinert, M., Thompson, K., Cooper, M.J. & Uzarski, D.G. (2012) Water level fluctuation and sediment-water nutrient exchange in Great Lakes coastal wetlands. Journal of Great Lakes Research, 38, 766-775.

Van Der Valk, A.G. (1981) Succession in wetlands: A Gleasonian approach. Ecology, 62, 688-696.

Van Der Valk, A.G. (2005) Water-level fluctuations in North American prairie wetlands. Hydrobiologia, 539, 171-188.

Van Geest, G.J., Coops, H., Roijackers, R.M.M., Buijse, A.D. & Scheffer, M. (2005) Succession of aquatic vegetation driven by reduced water-level fluctuations in floodplain lakes. Journal of Applied Ecology, 42, 251-260.

Van Wichelen, J., Declerck, S., Muylaert, K., Hoste, I., Geenens, V., Vandekerkhove, J., Michels, E., De Pauw, N., Hoffmann, M., De Meester, L. & Vyverman, W. (2007) The importance of drawdown and sediment removal for the restoration of the eutrophied shallow Lake Kraenepoel (Belgium). Hydrobiologia, 584, 291-303.

Van Wijck, C. & Degroot, C.J. (1993) The impact of disiccation of a frewater marsh (Garcine Nord, Camargue, France) on sediment water interactions. 2. The submerged macrophyte vegetation. Hydrobiologia, 252, 95-103.

Vermaat, J.E. & Sand-Jensen, K. (1987) Survival, metabolism and growth of Ulva lactuca under winter conditions - a laboratory study of bottlenecks in the life-cycle. Marine Biology, 95, 55-61.

Wetzel, R.G. (2001) Limnology: Lake and River Ecosystems, Elsevier, San Diego, California.

White, M.S., Xenopoulos, M.A., Hogsden, K., Metcalfe, R.A. & Dillon, P.J. (2008) Natural lake level fluctuation and associated concordance with water quality and aquatic communities within small lakes of the Laurentian Great Lakes region. Hydrobiologia, 613, 21-31.

 

101

 

102

103

 

104

66 URT 37:2 • maj 2013

Kransnålalgers udbredelseNår man graver nye vandhuller, dukker der ofte kransnålalger op. Kransnålalgernes sporer er robuste og spredes let med fugle. Måske kan de også overleve læn­ ge i jorden og spire frem, hvis der genetableres vandhuller på steder, hvor der engang tidligere har stået vand (Hutchinson 1975). Kransnålalgerne domine­rer vegetationen i de første år af vandhullets liv for senere at blive udkonkurreret og forsvinde, når rørsumpen etable res, eller højt­voksne undervands planter tager over. Kransnålalgerne hænger dog ved i flere årtier i kalkrige,

næringsfattige vandhuller, som vi blandt andet mø der i Vestskoven og i grus­ og kalk grave. Men krans nålalger er små, langsomt voksende arter som på sigt ofte udkonkurreres af kraftige blom­sterplanter.

Vi finder flest arter af krans­nålalger, især af hovedslægten Chara i klarvandede kalkrige per­manente søer. I kalkfattige søer med lav pH vokser færre arter, og de tilhører hovedsagelig slæg­ten Nitella. Kransnålalgerne vok­ser dels på lavt vand som en un­derskov mellem ret åbne bestande af rørsumps­, flyde­blads­ og undervandsplanter,

dels som en lav mere homogen vegetation på dybt vand, hvor lyset er for svagt til, at de rodfæ­stede blomsterplanter kan klare sig. Den kalkrige Furesø var i sine klar vandede velmagtsdage først i 1900­tallet en præ mie lo­kalitet med ti arter af kransnål­al ger, men eutrofieringen udryd­dede dem alle i en periode i 1960­80’erne. I dag er Maribo Søn dersø, Nors Sø og Magle Sø ved Brorfelde blandt de bedste kransnålalgelokaliteter med 6­8 arter. I mere end 90 % af alle danske lokaliteter finder vi imid­lertid blot 1­3 arter af krans nål­alger, og mere end halvdelen af vand områderne rummer kun en enkelt art. Nu om dage hører det også til de absolutte sjæl den­he der, at kransnålalgerne danner tætte bestande i søerne.

I beskyttede indre fjorde så­som Dybsø Fjord, Grund Fjord, Odense Fjord og Præstø Fjord voksede frem til 1950’erne både salttålende ferskvandsarter og egentlige brakvandsarter i tætte bestande, før eutrofiering og op­blomstring af planktonalger og

Kransnålalger rummer mange truede arterLars Baastrup-Spohr, Jeppe Dahl-Nielsen & Kaj Sand-Jensen

Kransnålalger har ry for at være svære at artsbestemme, vokse dybt og derfor kræve en kasterive og nogen gange en båd at indsamle. Derfor er de ofte blevet overset siden Sigurd Olsens klassiske studier (1944). Men nu har vi samlet informationer for de seneste årtier, som viser, at et flertal af arterne er enten sjældne, truede eller sandsyn­ligvis uddøde. Et forslag til rødliste understreger behovet for at for­bed re kransnålalgernes livsbetingelser og holde skarpt øje med udviklingen

Billede 1. Bugtet Glanstråd (Nitella flexilis) er en af de få arter af kransnålalger der er gået frem i det seneste år-hundrede, i både Danmark og andre vesteuropæiske lande. Arten var tidligere mest kendt fra det kolde bundvand i kalkfattige søer, men er nu vidt udbredt også i næringsrige søer og damme. Foto: Jens Christian Schou (biopix.dk)

105

URT 37:2• maj 2013 67

forureningsalger kvalte dem. Sådanne stjernelokaliteter for kransnålalger er forsvundet over alt både i de danske far van­ de og ved Tysklands og Polens Østersøkyster. Længere oppe i Østersøen langs de svenske og baltiske kyster, hvor eutrofierin­gen er mere beskeden, er der fortsat områder med store fore­komster af kransnålalger. Samlet set udgør kransnålalgerne imid­lertid én af de mest sjældne og udrydningstruede organisme­grupper i Europa.

Arternes relative hyppig­hed før og nuFor at kunne opgøre kransnålal­gernes status har vi sammenstil­let alle informationer om deres forekomst ifølge amternes og mil jøcentrenes optegnelser, vi­denskabelige studier og dygtige botanikeres observationer. End­videre har vi (JDN) selv besøgt et stort antal lokaliteter. I vurderin­gen af arternes nuværende rela­tive hyppighed tager vi udgangs­punkt i miljøovervågningens data, da disse er indsamlet efter den samme metodik. I alt indgår 399 artsobservationer fordelt på 226 lokaliteter i vores sammen­stilling, mens Sigurd Olsens data rummede 243 artsfund fordelt på 124 lokaliteter. Der voksede med sikkerhed 25 arter af krans­nålalger i Danmark for 70 år si­den, mens der i dag vokser 23 arter.

Arternes hyppighed før og nu er signifikant korrelerede (Fig. 1). Arternes hyppighed for 70 år si­den kan endvidere forklare 75 % af variationen i arternes hyppig­hed i dag. De mest alminde lige arter i begge studier er Chara glo­bularis, Chara vulgaris og Chara a spera. Nitella flexilis er hyppig (28.2 % af lokaliteterne) i det

nye datasæt, men faktisk ret få­tallig (3,5%) i det gamle datasæt. Det kan have betydning, at dybe klarvandede kalkfattige søer, som er et yndet voksested for Nitella flexilis, var underrepræsen­teret i Olsens undersøgelse, men det forklarer ikke alene den me­get store forskel. Lige som C. glo­bularis og C. vulgaris har N. flexilis nemlig vist sig at kunne trives under en bred vifte af næ rings­stofkoncentrationer, herunder meget eutrofe forhold. De tre nævnte arter er således også gået frem og er dominerende i andre vesteuropæiske lande med stor næringsstofpåvirkning, såsom Holland.

Både i den gamle og den nye sammenstilling optræder mange arter med en lav hyppighed, og de må derfor betegnes som sjældne og truede og dermed som kandidater til rødlisten. I Olsens kvantitative data optræ­der 8 arter således med en hyp­pighed på mellem 0,9 og 4% på lokaliteterne og yderligere 2 ar­ter er registreret som sjældne, mens 12 arter i det nye datasæt optræder med en hyppighed på mellem 0,8 og 4,2%. Otte af de fåtallige arter er fælles for de to sammenstillinger. For arterne med primær forekomst i brak­vand er de historiske data ikke lige så fyldestgørende, men op­teg nelser fra 1980’erne og frem viser, at flere arter er sjældne og i tilbagegang i danske farvande og også i resten af Østersøen.

Forsvundne og nye arterForsvundne arter Fem af Sigurd Olsens sjældne arter er ikke genfundet ved de seneste undersøgelser, og fire af arterne er sandsynligvis forsvun­det fra Danmark.

Tolypella intricata er sjælden i

Norden, og var længe vurderet som uddød i Sverige, men er i de senere år genfundet på få lo­ka liteter. Arten er enårig og har typisk en meteorisk forekomst, hvor den pludselig duk ker op, ofte i nydannede udtørrende habitater. Den er fundet i et oversvømmet hjulspor i Sverige. Ifølge Olsen (1944) var arten ob­serveret på 6 lokaliteter før år 1900, hvorefter arten ikke er ble­vet observeret. De periodevis oversvømmede habitater, som ar ten foretrækker, er blevet meget sjældne herhjemme i det drænede danske landskab. Det er derfor overvejende sandsyn­ligt, at arten er nationalt uddød.

Chara filiformis er udbredt i Midt­ og Østeuropa og voksede indtil 1939 i Furesø. Arten er i dag i Skandinavien kun kendt fra kalksøen, Levrasøen i det østlige Skåne. Den har en sære­gen morfologi og risikoen for fejl bestemmelse er derfor lille.

Chara horrida var tidligere hyp­pig i farvandene omkring Fyn og Nordtyskland, men her er un­dervandsvegetationen gået vold­som tilbage. I nyere tid (1990­

Fig. 1. Procentvis forekomst af arter i studier for 70 år siden (x-ak-sen, Olsen 1944) og i dag (y-aksen). Tallene er afbildet som logaritmen til den procentvise forekomst plus 1, så arter med nul tilstedeværelse har logaritmeværdien 0 i diagrammet.

106

68 URT 37:2 • maj 2013

2003) er arten kun kendt fra det indre af Østersøen og fra norske fjorde. Ifølge de nyeste DNA studier kan arten ikke skelnes genetisk fra C. baltica, ligesom overgangsformer mellem C. hor­rida og C. baltica også kendes fra Østersøen (Boegle et al. 2010). Uanset om C. horrida og C. baltica er separate arter, er ingen af dem observeret i C. horrida’s tid­ligere kerneområde i det sydfyn­ske øhav i nyere tid. Vi kan ikke med stor sikkerhed konkludere at arten er forsvundet.

Nitella gracilis er sjælden i vores nabolande Tyskland, Norge og Sverige og er tidligere kun fun­det på én lokalitet herhjemme (Olsen 1944). Nitella gracilis kan ligne N. mucronata og kan kun med sikkerhed adskilles fra denne ved mikroskopi af spor­erne. I Danmark er der i nyere tid kun få fund af N. mucronata (JDN og M. Holmen), og deres identifikation er bekræftet af uafhængige eksperter. Derfor er Nitella gracilis sandsynligvis for­svundet fra Danmark.

Nitella opaca er kun genfundet meget få gange Danmark i de seneste årtier. Den er ikke truet i Sverige og Norge, men er truet i Nordtyskland. Vegetativt kan ar­ten ikke adskilles fra Nitella flexi­lis. Men da N. flexilis er almind­elig, kan N. opaca fortsat skjule sig bag de mange ikke artsbe­stemte fund af Nitella.

Tabel 1. Forekomsten af kransnålalger i Danmark angivet som antallet af voksesteder. Data er fra de tidligere amter og miljøcentres overvågning af søer med observationer fra ca. 1989 til 2011 i marine områder og 1995-2011 i ferskvand. De marine registreringer er suppleret med observationer fra Statens Naturhistoriske Museum. Fugleognatur.dk er offentlig brugersty-ret naturdatabase med observationer af kransnålalger fra 2006-2012. Data fra de forskellige kilder ikke nødvedigvis adderes, da det ikke kan ude-lukkes, at nogle voksesteder kan gå igen i de forskellige kilder, og det totale voksesteder er derfor angivet som det maksimale antal observerede lokal-iteter. Før og nu henviser en sammenligning af arternes procentvise fore-komst i ca. 1940 og til ca. 2010 baseret på tal fra miljøovervågningen og Olsen (1944) og beregnet af Baastrup-Spohr et al. (2013). Den enkelte arts rødliste status er primært baseret på antallet voksesteder samt den tidslige udvikling i den procentvise forekomst (se Wind og Pihl, 2010).

Art Miljø­over­våg­ning

Fugleog natur.dk

Total antal lok.

Før og nu Foreslået rødliste status

Chara aspera 59 5 59 11.4­16.5 LC

Chara baltica 25** 25 NT

Chara canescens 27** 27 NT

Chara connivens 8* 1 8 0.0­2.4 VU

Chara contraria 21 5 21 12.3­10.0 NT

Chara denudata 3* 3 2.6­1.2 EN

Chara filiformis 0 0 0.9­0 RE

Chara globularis 106 25 106 59.6­56.5 LC

Chara hispida 12 15 15 16.7­5.9 NT

Chara horrida 0 0 RE

Chara intermedia 1 2 2 11.4­0.0 EN

Chara polyacantha 1 1 1 1.8­0.0 CR

Chara rudis 1 1 7.0­0.6 CR

Chara tomentosa 9 9 7.9­4.1 VU

Chara virgata 12 5 12 N/A NT

Chara vulgaris 69 18 69 26.3­32.4 LC

Lamprothamnium papulosum

6 11 EN

Nitella capillaris 2 1 2 En

Nitella flexilis 48 6 48 3.5­28.2 LC

Nitella gracilis 0 0 0.9­0.0 RE

Nitella mucronata 2* 2 2 0.9­0.0 EN

Nitella opaca 3* 3 5.3­0.0 EN

Nitella translucens 7 7 2.6­4.1 NT

Nittelopsis obtusa 14 1 14 6.1­7.6 NT

Tolypella glomerata 7* 7 2.6­1.2 VU

Tolypella intricata 0 0 RE

Tolypella nidifica 32** 32 NT

*antallet er en kombination af pri-vate observatørers observationer og voksesteder registreret i miljøovervågningen. **antallet er en kombination af ob-servationer fra Statens Naturhistor-iske Museum og voksesteder regis-treret i miljøovervågningen i perioden (1997-nu).

107

URT 37:2• maj 2013 69

Nye arterTo nye arter optræder i de sene­ste opgørelser, én af dem anser vi for nyindvandrede. Chara con­nivens lever primært i lav­salint brakvand og alkalint ferskvand. Arten er sjælden i det meste af Østersøen, men er dog fundet i stort antal ved Estland. I Dan­mark er den fundet på mindst 6 lokaliteter af forskellige perso­ner.

Chara virgata optræder i de nye undersøgelser, men er i Olsens studier inkluderet som et syno­nym af Chara globularis (før kaldet C. fragilis). C. virgata er derfor med stor sandsynlighed ikke ny for Danmark, men har blot væ­ret inkluderet i C. globularis, som den ligner meget.

Et revideret nationalt arts­tal og forslag til rødlisteDa Chara virgata ikke er ny for Danmark og hvis ikke Nitella opa ca er forsvundet, skal vi regne med 26 arter af krans nål alger på Sigurd Olsens tid og 23 arter i dag. Det ligger tæt på arts tallet i Irland og England (25) og Nor ge (26), mens Sverige (31) og Tyskland (35) har flere registre­rede arter.

Som bekendt rummer en rød­liste seks kategorier af arter, som vurderes som: ikke truede (LC), næsten truede (NT), sårbare (VU), moderat truede (EN), kri­tisk truede (CR) eller uddøde (RE). En arts placering på listen kan begrundes ud fra flere inter­natonalt fastlagte kriterier, pri­

mært med fokus på tilbagegang af artens geografiske udbredel­sesområde eller bestands stør rel­sen. En art optræder altså som regel ikke på rødlisten, hvis den ”bare” er sjælden, men der bli­ver selvfølgelig lagt vægt på sjæld ne arter i tilbagegang. Listerne udarbejdes af Dansk Center for Miljø og Energi (DCE) og Naturstyrelsen, men det er ikke sket for kransnålal­ger nes vedkommende, sikkert pga. manglende kendskab til arternes hyppighed og udvikling.

Da vi har samlet den til gæn ge­lige viden omkring denne over­sete gruppe, foreslår vi her en rødlistestatus for de danske arter af kransnålalger (Tabel 1). Vur­deringen er primært begrundet i

Billede 2. Den karakteriske tyk kransnål (Chara tomentosa) kendes på sit brutale udtryk, snoede bark og de ofte rødfarvede skud. Arten var tidligere ret almindelig i klarvandede kalkrige søer, men den er gået tilbage og findes nu kun i få danske søer. Foto: Jens Christian Schou (biopix.dk).

108

70 URT 37:2 • maj 2013

antallet af kendte vokseste der for den enkelt art samt vores kvantitative undersøgelse af hi­storiske ændringer for arten (Baastrup­Spohr et al. 2013). Antallet af voksesteder er vurderet til at være det maksi­male antal lokaliteter observeret i en af de tre benyttede kilder (Tabel 1). For de fleste arter er det højeste antal lokaliteter be­skrevet i miljøovervågningens data, men især for arter med hovedforekomst i brakvand er der flere registreringer i Schu­bert & Blindow (2003). For en række ret sjældne arter ofte til­knyttet småvande er det højeste antal voksesteder fundet på fug­leognatur.dk, hvilket sandsynlig­vis skyldes, at disse habitater ofte er små lavvandede lokali te ter, der ikke inventeres i overvåg­nings programmerne, men er tilgængelige for private entusia­ster.

Kun 4 af 26 arter (15%) vur­deres at have en gunstig status mens de resterende 85 % er tru­ede i mere eller mindre grad. I gruppen af truede arter finder vi de nævnte formodede regio nalt uddøde arter (4), de to kritisk truede Chara polyacantha og Chara rudis som hver kun har et obser­veret voksested i Danmark, og som begge har været i til ba ge­gang siden Olsens studier i 1940’erne. Den resterende gruppe (14 arter) har ligeledes et meget begrænset antal vokse­steder og de fleste er gået tilbage i løbet af det 20. århundrede.

Kransnålalgerne som gruppe er derfor stærkt truede. Vi håber derfor, at vi med dette rødliste­forslag kan øge opmærksomhe­den omkring denne ofte oversete gruppe, for at forbedre vurderin­gen af arternes hyppighed og bedre sikre deres levesteder.

LitteraturBaastrup­Spohr, L., Iversen, L.L., Dahl­

Nielsen, J & Sand­Jensen, K. 2013. Seventy years of change in abun­dance of Danish charophytes. Fresh­water Biology. Under trykning.

Boegle, M. G., Schneider, S. C., Melzer, A. & Schubert,H. 2010. Distinguish­ing Chara baltica, C. horrida and C. lilje­bladii – conflicting results from analy­sis of morphology and genetics. Charophytes 2: 53­58.

Hutchinson, G.E. 1975. A treatise on limnology. Limnological botany (Vol 3). Wiley.

Olsen, S. 1944. Danish Charaphyta. Chronological, ecological and biologi­cal investigation. Ejnar Munksgaard.

Schubert, H. & Blindow, I. 2003. Charo­phytes of the Baltic Sea. ARg Gartner

Wind, P. & Pihl, S. (red.) 2010. Den dan­ske rødliste. Dansk Center for Miljø og Energi (DCE), Aarhus Universitet.

Forfatternes adresse: Københavns Universitet, Ferskvandsbi­ologisk Laboratorium, Helsingørsgade 51, 3400 Hillerød

Botanisk notitser…

Rettelse til URT nr. 1, 2013I forbindelse med ’Årets fund 2012 fra hele Danmark’ er der i ’Øserne øst for Storebælt’ ind­sneget en forkert stedangivelse. Sølv­elefantgræs (Miscanthus sac­chariflorus) vokser ikke på østsi­den af Skovledsmosen i Bal le­rup, men på østsiden af den lille sø der ligger nordøst for Skov­ledsmosen. Se også på http://map.krak.dk/m/nm98J.

Havens RødderHavens Rødder er en netværks­gruppe, hvis formål det er at ind drage Botanisk Have i den gymnasiale naturvidenskabelige undervisning, specielt i biologi men også i tværfaglige projekter. Vi vil gerne slå et slag for, at bo­taniske emner og problemstillin­

ger udgør en større og vigtigere del af den gymnasiale undervis­ning end de gør i dag. Det er og så i denne sammenhæng vig­tigt at pege på Botanisk Haves potentialer.

Vi er ca. 20 personer i grup­pen. Den består af • forskere fra Botanisk Have,

Bio logisk institut, LIFE og Pharma både nuværende og tidligere,

• gymnasielærere med fagene biologi, kemi, geografi og matematik,

• andre der interesser sig for Botanisk Have.

Læs evt. mere om formålet med Røddernes virke på hjemmesi­den på brughaven.dk

Fra hjemmesiden kan man også downloade hæftet Brug Bo tanisk Have i Undervisnin­gen som indeholder artikler

elever kan læse før et besøg i Haven. Det kan i øvrigt købes i Havens butik for 65 kr.

Vi mangler botanisk vidende og interesserede personer, der har lyst til at assistere gymna sie­klasser, ­grupper eller ­opga ve­skri vere ved besøg i haven og ved arbejde med forskellige pro­jekter, der indeholder botaniske problemstillinger. Vi har også brug for faglige oplæg, når rød­derne mødes.

Kontakt os, hvis I er interes se­ret og fortæl evt. samtidig, hvil­ket emne/emner I gerne vil give oplæg om. Så sætter vi jer i for­bindelse med dem, der har brug for netop jeres specielle eksper­tise.

Kontakt: Dorte Hammelev, dorte@centrum.dk

109

 

110

111

 

112

SVENSK BOTANISK TIDSKRIFT 104:1 (2010) 23

Extrema miljöer pressar växterna till det yttersta och skapar ofta tydliga fördelnings­mönster mellan arterna beroende på deras olika härdighet och konkurrensförmåga. Så förhåller det sig också i övergången från uttorkade kalk­stenshällar via tillfälliga till permanenta vatten­samlingar i ett kalkbrott på Räpplinge alvar. En forskargrupp från Köpenhamns universitet besökte Öland i maj 2009 för att undersöka den speciella miljön.

KAJ SAND-JENSEN, LARS BAASTRUP-SPOHR,

ANDERS WINKEL, CLAUS LINDSKOV MØLLER,

JENS BORUM, KLAUS PETER BRODERSEN,

TORBJÖRN LINDELL & PETER ANTON STÆHR

Vad som vid första påseendet verkade vara ett öde och ointressant landskap i ett stenbrott på Räpplinge alvar bjuder i själva verket på

ett fascinerande utbud av växter med utomordent-lig anpassningsförmåga. Inom några få meter skif-tar miljön från vänlig och tillmötesgående med

god vattentillgång, ordentligt fäste för växternas rötter och små temperatursvängningar, till den mest fientliga, där en sjö förvandlas till en öken på några få dagar och där temperaturen varierar från bitande kyla till stekhet solvärme.

Fascinationen förstärks av att man här finner både den utpräglade torrmarksväxten vit fet-knopp, som lever av nästan ingenting under torr-perioder och överlever vinterns översvämningar, och vattenväxten skyfallsalg, som överlever torkan genom att ”spela död” men omedelbart återupp-står från de döda när vattnet kommer tillbaka.

Miljögradienter och växtfördelningarTydliga fördelningsmönster mellan växtarter finner man längs gradienter för resurser som ljus, fuktighet och närsalter (Vestergaard 2007). Gradienter av andra miljöfaktorer som vind, tem-peratur, salthalt och pH skapar också tydliga för-delningar (Tyler 1971, Crain m.fl. 2004). I regel förekommer många arter längs gradienten, och arternas utbredningar går ofta gradvis in i varan-

Ett kalkbrott på Ölands alvar – en stenöken med knivskarpa miljögränser

Figur 1. Räpplinge stenbrott. Mellan de 5–10 cm tjocka skikten av hård kalksten finns mer lättvittrat material. Foto: Lars Baastrup-Spohr.The lime quarry at Räpplinge alvar. Profile of 5–10 cm thick layers of hard limestone with thinner and more easily disin-tegrating layers inbetween.

113

SAND-JENSEN m.fl.

24 SVENSK BOTANISK TIDSKRIFT 104:1 (2010)

dras. Negativa eller positiva interaktioner mellan olika arter påverkar också fördelningen (Maestre m.fl. 2009), och då i synnerhet konkurrens.

Om det finns få arter blir gränserna särskilt påtagliga och om flera av miljöfaktorerna för-stärker varandra blir växlingen mellan arterna ännu tydligare. Man kan till exempel se en tydlig zonering i sjöar med ökande djup där hela bälten med enartsbestånd avlöser varandra beroende på skillnader i belysningskrav, tolerans mot olika vat-tendjup och konkurrensförmåga (Sand-Jensen & Søndergaard 1997).

Ett liknande fenomen kan man se i avrinnings-områden från brunkolsbrott, där extremt surt vatten (pH 2–4) med höga koncentrationer av metaller bara medger att en enda blomväxt förmår överleva (löktåg Juncus bulbosus), men där fler

arter ansluter i takt med att pH-värdet stiger och metallkoncentrationen sjunker när allt renare vatten från omgivningen tillförs (Sand-Jensen & Rasmussen 1977).

I miljöer där det finns få arter och där gradien-terna är påtagliga och fördelningen mellan arterna tydlig, blir det möjligt att testa om det är speciella fysiologiska och ekologiska anpassningar hos arterna eller om det är skillnader i konkurrensför-måga som förklarar varför de växer där de växer längs gradienten.

En sådan mycket lämplig situation hittar vi längs gradienten från land till vatten i stenbrottet vid Räpplinge alvar på Öland. Här ändrar sig vat-ten- och näringstillgången snabbt och påtagligt och när förhållandena är som mest extrema finner man endast några få arter.

Figur 2. De uttorkade kalkhällarna på Räpplinge alvar täcks av vit fetknopp Sedum album och svarta intor-kade skorpor av skyfallsalg Nostoc commune. Foto: Jens Borum.Dried out limestone slabs covered by reddish Sedum album and black, dry crusts of Nostoc commune.

114

STENÖKEN

SVENSK BOTANISK TIDSKRIFT 104:1 (2010) 25

En varierande miljö i stenbrottetDen hårda kalkstenen är bruten i horisontella skikt (figur 1). Mellan de 5–10 cm tjocka kalk-flisorna finns mer lättvittrat material som till-sammans med sönderbrutna kalkflisor samlats i avfallshögar av 2–3 meters höjd. Dessa avfallshö-gar är bevuxna med buskar och ett stort antal olika örter, inklusive johannesnycklar Orchis militaris och ängsnycklar Dactylorhiza incarnata. Denna vegetation har vi inte närmare undersökt.Däremot har vi undersökt de endast några få cen-timeter höga övergångszonerna mellan kalkstens-ytorna, där damm och jordpartiklar har samlats och så småningom bildat ett 1–4 cm tjockt jord-lager, som ger växterna rotfäste. Jorden är något fuktig vilket gynnar växtlivet, som kan omfatta mer än trettio olika arter.

På de helt plana kalkstensytorna växer endast vit fetknopp Sedum album tillsammans med gul fetknopp S. acre och tåliga mossor (figur 2). Här finns endast ett mycket tunt jordlager som

inte kan bevara någon fuktighet, så här krävs en extrem anpassning till frost på vintern, stekande hetta på sommaren, bitande vind, samt växlingar mellan översvämning och torka för att kunna överleva.

I områden som har stått under vatten från oktober till mars är ytorna täckta med intorkade exemplar av cyanobakterien skyfallsalg Nostoc commune, som bildar mycket anmärkningsvärda kolonier (figur 3). Under vatten liknar de 3–5 mm tjockt, mörkgrönt vingummi. På 1–2 cm djup bildar cyanobakterien tunna blåsor, som sticker upp över vattenytan, men som faller samman vid uttorkning. Sådana grunda vattensamlingar finns nedanför avfallshögarna och från dessa sipprar vatten ut över den flata kalkstenshällen. I dessa kan det fortfarande finnas vatten ännu i slutet av maj. Men bara några få veckor senare är vatten-samlingarna helt uttorkade och skyfallsalgerna har bildat stora, tunna skorpor. I händelse av kraf-tigt regn kan de på nytt omedelbart återskapa de

Figur 3. Skyfallsalg Nostoc commune torkar in och bildar en svart inaktiv skorpa i luft (vänster). Den suger upp vatten och blir aktiv på några få sekunder när den får tillgång till vatten (mitten). Efter tre dagar i vatten (höger) har algen blivit ännu tjockare, grönare och tre gånger mera aktiv jämfört med förhållan-dena ett par timmar efter tillförseln av vatten. Foto: Jens Borum.Dry crusts of Nostoc commune (left) can rapidly absorb water, turn metabolically active (middle) and gradu-ally become even thicker and greener within a couple of days (right).

115

SAND-JENSEN m.fl.

26 SVENSK BOTANISK TIDSKRIFT 104:1 (2010)

aktiva, geléartade kolonierna – därav förmodligen namnet skyfallsalg.

På lite djupare vatten på kalkhällen finns till-fälliga vattensamlingar som ännu i slutet av maj har ett vattendjup på några få centmeter, men som torkar ut i juni–juli. På några ställen har kalkste-nen brutits lite djupare och här kan man finna vattensamlingar med ett djup av 20–40 cm, vilka är vattenfyllda året om (figur 4). Vid besöket i maj var den uppmätta alkaliniteten högst i de perma-nenta vattensamlingarna, men dessa värden var endast 20–40 procent lägre i de grundare vatten-samlingarna (figur 5). Nivåerna motsvarar medel-hårt vatten. Ingen av vattensamlingarna innehöll surt vatten liknande regnvatten. När det regnar löses således kalkstoftet på hällarna upp och bildar alkaliskt vatten.

De grunda vattensamlingarna domineras av skyfallsalger samt av mattor av trådalger i de delar som har sediment. Fram på eftermiddagen kan här uppmätas ett pH på närmare 10, vilket innebär att koldioxidkoncentrationerna är försumbara och hundra gånger lägre än i luften. Det höga pH-värdet och underskottet av koldioxid beror på den intensiva fotosyntesen samt växternas förmåga att aktivt utnyttja vätekarbonat (HCO3

–). pH nådde inte riktigt upp till samma nivå (ca 9,3) i de djupare vattensamlingarna. Här dämpas pH-vari-

Figur 4. Några av artikelförfattarna står vid en av de permanenta vattensamlingarna i stenbrottet på Räpplinge alvar. Foto: Jens Borum.Permanent pond in the stone quarry visited by some of the authors.

Figur 5. Miljöförhållanden och artsammansättning i tio områden längs en land-vattengradient. Område 1: Zon mellan kalkstenshällar med jordbildning och viss fuktighet. Område 2: Nästan kal kalkstenshäll som var uttorkad för 1–2 månader sedan. Område 3–8: Sex tillfällliga men ännu vattenfyllda vatten-samlingar med stigande djup. Område 9–10: Två djupare, permanenta vattensamlingar.

Land-vattengradienten kan beskrivas utifrån de torra områdenas belägenhet en bit ovanför vat-tenytan och av vattensamlingarnas ökande djup (panel 1). Jord- eller sedimenttjockleken är mycket låg på den öppna kalkstenshällen (område 2) och i de grunda vattensamlingarna (3–8) men högre i övergången mellan kalkstenshällar (1) och i de permanenta vattensamlingarna (9–10, panel 1). Antalet amfibiska och akvatiska vattenväxter stiger längs gradienten (panel 2). Alkaliniteten är medel-hög och pH-värdet är mycket högt på grund av intensiv fotosyntes under dagen (panel 3). Changes in elevation and water depth (open circles) and soil/sediment thickness in the transition from cracks in the limestone (1), over almost naked limestone plates (2) and temporary ponds (3–8) to permanent ponds (9–10; panel 1). Also shown are numbers of amphibious (open columns) and sub-merged species (dark columns; panel 2) as well as pH and alkalinity in the ponds at noon (panel 3).

Område

pHA

ntal

arte

r

Amfibiska arter

Höj

d (c

m)

Jorddjup (mm

)A

lkalinitet (meq l -1)

Undervattensarter

116

STENÖKEN

SVENSK BOTANISK TIDSKRIFT 104:1 (2010) 27

ationen över dygnet av den större vattenmassan och av de här dominerande kransalgernas massiva kalkutfällning under sin fotosyntes vilket motver-kar en höjning av pH-värdet (se nedan).

Vattnet har ett mycket lågt näringsinnehåll i vattensamlingarna. Det gäller både ammonium (mindre än 40 mikrogram kväve per liter) och, i ännu högre grad, innehållet av oorganiskt upplöst fosfat (mindre än 4 mikrogram fosfor per liter i

flertalet vattensamlingar). Växter som växer på sedimentet eller har rötter eller rhizoider, kan i gengäld utnyttja detta sediment som en något bättre näringskälla. Förutom att vara anpassade till extrem variation i temperatur, vattentillgång och pH, måste växterna dessutom tåla varaktig fosforbrist, och under vatten kunna tåla ett högt pH, mycket låg koldioxidkoncentration samt kunna utnyttja vätekarbonat för sin fotosyntes. Landväxter kan inte utnyttja vätekarbonat och därför utgör vatten med högt pH ett hinder för deras överlevnad.

Olika växter i vattensamlingarnaSkyfallsalg i små uttorkade vattensamlingarSkyfallsalg och vit fetknopp utgör de domineran-de arterna på kalkstenshällarna. De torra kalkhäl-larna – som en månad tidigare stod under vatten

– har nu i maj svarta, intorkade skorpor av skyfall-salger mellan blodröda partier med vit fetknopp (figur 6, område 2). Det är tänkbart, men inte närmare undersökt, om skyfallsalgen tack vare sin förmåga att binda atmosfäriskt kväve gynnar etableringen av vit och gul fetknopp på den nakna kalkstenen. Det skulle i så fall vara ett exempel på ett gynnsamt samspel i ett område som präglas av brist på resurser.

I de mest temporära vattensamlingarna på den nakna kalkstenshällen kan skyfallsalger täcka 75 procent av botten, medan resten består av vit fet-knopp och naken kalksten (figur 6, område 3).Om vattensamlingarna är lite djupare och har ett några millimeter tjockt sediment etablerar sig kransalger, medan man i strandkanten återfinner spridda exemplar av kärrkavle Alopecurus genicula-tus och krypven Agrostis stolonifera.

Vattenmöja i större uttorkade vattensamlingarI djupare, men ändå uttorkade vattensamlingar med ett 1–2 cm tjockt sediment ovanpå kalk-hällen växer vattenmöja Ranunculus aquatilis, hårsärv Zannichellia palustris och en filt av tråd-alger på sedimentet medan skyfallsalger återfinns på den nakna kalkstenshällen vid de allra lägsta vattennivåerna (figur 6, område 4–7). På något djupare vatten dominerar kransalger. I de delar av vattensamlingarna som nyligen har blivit torr-

Figur 6. Procentuell täckning av olika växter längs land-vattengradienten. Områdesindelning, se figur 5.

Övre panelen: Landväxter (hela kolumnen, inkl. bar jord i grått), vit fetknopp Sedum album (streck-at), och övriga landväxter (vitt). Mellersta pane-len: Amfibiska arter (hela kolumnen), skyfallsalg (grått), vattenmöja (streckat) och övriga amfibiska arter (vitt). Nedre panelen: Undervattensarter (hela kolumnen), trådalger (streckat), kransalger (grått) och övriga arter (vitt).Cover (%) of terrestrial plants (upper panel), amphibious plants (middle panel) and submerged plants (lower panel) along the air-water gradient.

Amfibieväxter

Undervattensväxter

Landväxter och bar jord

Täck

ning

sgra

d (%

)

Område

Täck

ning

sgra

d (%

)Tä

ckni

ngsg

rad

(%)

117

SAND-JENSEN m.fl.

28 SVENSK BOTANISK TIDSKRIFT 104:1 (2010)

lagda, är kransalgerna döda men har satt oosporer som kan gro när vattnet återkommer.

Vattenmöja tål inte heller att torka ut, men tack vare fuktigheten i sedimentet kan de klara sig lite längre här än på den bara kalkhällen, där endast skyfallsalgen kan överleva.

Vattenmöja kan bilda både undervattens-, flyt- och övervattensblad men arten klarar sig först när det finns ett par centimeter sediment för att säkra näringstillgången (och vattentillgången på land), och landformen växer endast mellan stenar som skyddar mot vindens uttorkande effekt.

Kransalger i de djupaste vattensamlingarnaI de permanenta vattensamlingarna är vattnet djupare och sedimenten upp till 6 cm tjocka. Här kan kransalger täcka ungefär 80 procent av bott-nen (figur 6, område 9 och 10). Mellan de täta bestånden av kransalger växer enstaka exemplar av axslinga Myriophyllum spicatum, krusnate Potamogeton crispus och gäddnate P. natans. Sedimenten bildas eftersom vattensamlingarna fungerar som en fälla för jordpartiklar som kom-mer med vatten och vind från omgivningen. I sedimenten ansamlas också kalk och organiskt material som producerats av växterna.

På grunda delar på bottnar som saknar krans-alger finner man trådalger, vattenmöja, hår särv och, om man har tur, ävjebrodd Limosella aqua-tica. Här förekommer också en del halvgräs och andra växter som skjuter upp blad över vattenytan.

Arternas ekofysiologi och konkurrens Vit fetknopp har en enastående förmåga att växa på torra, nakna kalkhällar. Den är föga närings-krävande, har god vattenhållande förmåga och kan vid uttorkning ytterligare minska avdunst-ningen genom sin så kallade CAM-metabolism. Denna innebär att intaget av koldioxid huvud-sakligen sker om natten, medan klyvöppningarna är stängda på dagen. Fetknoppsarterna kan vid god vattentillgång även använda den vanliga C3-fotosyntesen under dagtid (Schube & Kluge 1981). Om jordlagret är tunt, möter vit fetknopp obetydlig konkurrens och bildar under fuktiga förhållanden täta högröda mattor (figur 2), men om jordlagret är lite tjockare tappar arten i kon-kurrensförmåga mot bland annat olika gräs och halvgräs. Vit fetknopp kan överleva under vatten under kortare perioder men den kan inte utnyttja vattnets vätekarbonat för sin fotosyntes (tabell 1), och täcks den av vatten på sommaren dör den.

Tabell 1. Växternas förmåga till att utföra fotosyntes under vatten i tillslutna flaskor under 12–14 timmar mätt genom extraktion av oorganiskt kol från vattnet (% av DIC-innehållet), höjning av pH-värdet (slut-pH) och sänkning av koldioxidhalten. Några av landväxterna är testade först efter de fått stå i vatten under två timmar.Capability to extract inorganic carbon (DIC), elevate pH, and to reduce free CO2 during photosynthesis in closed bottle over 12–14 hours of photosynthesis.

Kolupptag (%) Slut-pH CO2 (µmol L–1)LandväxterVit fetknopp Sedum album, torr 1,1 8,17 19,1Vit fetknopp, fuktig 1,0 8,28 16,9Takmossa Syntrichia ruralis, torr 1,0 8,25 15,3Takmossa, fuktig 1,0 8,30 15,5

Amfibiska växterVattenmöja Ranunculus aquatilis, luftblad –0,6 7,96 31,8Vattenmöja, undervattensblad 48,3 9,84 0,08Skyfallsalg Nostoc commune, torr 43,1 9,43 0,65Skyfallsalg, vatten 28,7 9,30 0,98Skyfallsalg, vuxit i vatten 54,6 9,92 0,15

UndervattensväxterTaggsträfse Chara hispida 55,0 9,59 0,11Axslinga Myriophyllum spicatum 28,5 9,45 0,13

118

STENÖKEN

SVENSK BOTANISK TIDSKRIFT 104:1 (2010) 29

Takmossa Syntrichia ruralis blir inaktiv vid torka, men den överlever på de nakna kalkstens-hällarna tillsammans med fetknopparna och tre andra mossor: kalklockmossa Homalothecium lutescens, plyschmossa Ditrichum flexicaule och sandraggmossa Racomitrium canescens. Tillväxt och fotosyntes kräver regnvatten eller fuktig jord.

Under vatten överlever takmossan men det är inte klarlagt hur länge. Då den endast kan använ-da sig av koldioxid och inte vätekarbonat (tabell 1), växer den varken i tillfälliga eller permanenta vattensamlingar, där skyfallsalger, vattenmöja och kransalger pressar upp pH-värdet så högt – och därmed koldioxidinnehållet så långt ner – att mossorna inte kan ta upp någon koldioxid. Alla mossarna visar samma respons. Andra landväxter får inte heller tillräckligt med koldioxid om de täcks av vatten.

Takmossan och de övriga mossarna växer där-för enbart i ett snävt intervall av den studerade miljö gradienten, på torra till halvfuktiga och ytterst tunna jordar där de kan dra nytta av sin förmåga att tåla uttorkning. De upphör då med ämnesomsättningen men återupptar aktiviteten snabbt när de åter får vatten. Alla fyra mossorna bildar täta kuddar som likt svampar suger upp vatten och också förlorar vattnet långsamt, vilket medger en förlängd aktivitetsperiod.

Skyfallsalgen är i enastående hög grad anpassad till livet i tillfälliga vattensamlingar på de nakna kalkstenshällarna. I uttorkat tillstånd är den full-ständigt inaktiv men fortfarande vid liv. Efter mer än hundra år på ett herbarieark kan skyfallsalger åter väckas till liv om vatten tillförs. Inom en minut har kolonierna sugit upp så mycket vatten att deras vikt ökat åtta gånger, och inom samma tidsrymd startar respirationen (Scherer m.fl. 1984). Fotosyntesen följer omedelbart efter, och blir större än respirationen redan 5–10 minuter efter det att vatten tillförts. Inom loppet av några dagar ökar klorofyllinnehållet, bålen blir tjockare och ljusabsorption, fotosyntes och tillväxthastig-het stiger (figur 3).

Skyfallsalgen utnyttjar effektivt vätekarbonat i vatten (tabell 1). Den klarar sig också utan tillför-sel av nitrat och ammonium tack vare sin förmåga till kvävefixering. Skyfallsalgen är aktiv så länge

den får vatten, men när underlaget torkar ut går kolonierna snabbt samma väg, varvid fotosyntes och respiration upphör. Skyfallsalgen besitter såle-des en imponerande anpassning till perioder med omväxlande extrem torka och översvämning.

Vattenmöja dominerar i tillfälliga dammar med sediment av några få centimeters tjocklek (figur 6). Den har speciella förutsättningar för att kunna hålla sig aktiv på såväl land som i vatten, eftersom den kan bilda olika bladtyper, anpassade antingen till att fungera under vatten, på vatten-ytan eller ovanför denna (Nielsen 1993). Under-vattensbladen är hårformiga, flytbladen utgörs av helbräddade bladskivor medan över vattensbladen är formade som undervattensbladen, men är något tjockare som begränsar avdunstningen och försedda med klyvöppningar för att utnyttja luftens koldioxid. Undervattensbladen är väl anpassade till att utnyttja vätekarbonat för sin fotosyntes och klarar utan vidare ett pH-värde närmare 10 (tabell 1). Under blomningen utnytt-jas flytbladen som en plattform för att lyfta blom-morna upp i luften. Torkar vattensamlingarna ut, utvecklas övervattensbladen så att fotosyntes i luft kan ske medan undervattensbladen snabbt torkar ihop och dör. Vattenmöja klarar övergången mel-lan den torra och den våta miljön väl och bibehål-ler sin aktivitet i båda fallen tack vare sina olika blad.

Kransalgen taggsträfse Chara hispida domine-rar i de permanenta vattensamlingarna, och den finns också i de djupare delarna av de tillfälliga vattensamlingarna innan de torkar ut (figur 6). Taggsträfse utnyttjar effektivt vätekarbonat som kolkälla för sin fotosyntes (tabell 1) genom att i en aktiv process fälla ut kalk på cellytorna. För varje utfälld molekyl kalciumkarbonat frigörs en vätejon som förvandlar en annan vätekarbonatjon till fri koldioxid som används i fotosyntesen. Tack vare kalkutfällningen kan fotosyntesen fortsätta utan att pH-värdet blir extremt högt och hämmar vidare aktivitet.

Tre mästare i anpassningPå kalkstenshällarna vid Räpplinge alvar kan man således direkt se vilka arter som kan klara de mest extrema miljöförhållandena.

119

SAND-JENSEN m.fl.

30 SVENSK BOTANISK TIDSKRIFT 104:1 (2010)

En mästare i att överleva i denna extrema miljö är vit fetknopp, som upprätthåller sin aktivitet på land trots torka, medan dess fotosyntes går i stå under vatten på grund av brist på oorganiskt kol. Den andra är takmossan som är aktiv i fuktigt tillstånd på land. Den tredje är skyfallsalgen, som är aktiv både i fuktigt tillstånd på land samt under vatten. Alla tre arterna är endast några få centi-meter höga och är vatten- och näringstillgången mer riklig konkurreras de ut av en mera artrik och högväxt vegetation med större biomassa.

Man kan fråga sig hur vit fetknopp, takmossa och skyfallsalg får tillräckligt med närsalter ur regnvattnet och den hårda kalkhällen. Särskilt är det ett problem att få tag i fosfat som är hårt bun-det till kalkfosfater, men också järntillgången kan vara otillräcklig i kalkhaltiga jordar. Skyfallsalgen använder sig av aktiva jonpumpar för att kunna utnyttja vätekarbonat till fotosyntesen i vatten med högt pH, och den har som andra cyanobakte-rier sannolikt specifika proteiner i cellmembranen för at ta upp fosfat (Zhang m.fl. 2009). Det är också tänkbart att skyfallsalgen kan avge orga-niska syror för att upplösa kalkfosfater, något som man funnit hos andra kalkväxter (Ström m.fl. 2005), och den vita fetknoppens rötter samt tak-mossans rhizoider kan kanske göra något liknande. Det är dock den robusta byggnaden i kombination med överlevnadsförmågan som är huvudorsakerna till framgången för vit fetknopp, mossor och sky-fallsalgen i denna öländska stenöken.

• Tack till Carlsbergsfonden för ekonomiskt stöd.

Citerad litterarurCrain, C. M., Silliman, B. R., Bertness, S. L. & Bertness,

M. D. 2004. Physical and biological drivers of plant distribution across estuarine salinity gradients. – Ecology 85: 2539–2549.

Maestre, F. T., Callaway, R. M., Valladares, F. & Lortie, C. J. 2009. Refining the stress-gradient hypothesis for competion and facilitation in plant communi-ties. – J. Ecol. 97: 199–2005.

Nielsen, S. L. 1993. A comparison of aerial and sub-merged photosynthesis in some Danish amphibious plants. – Aquat. Bot. 45: 27–40.

Sand-Jensen, K. & Rasmussen, L. 1977. Macrophytes and chemistry of acidic streams from lignite mining area. – Bot. Tidsskrift 72: 105–111.

Sand-Jensen, K. & Søndergaard, M. 1997. Plants and environmental conditions in Danish Lobelia lakes.

– I: Sand-Jensen, K. & Pedersen, O. (red.), Fresh-water biology. Priorities and development in Danish research. Gad, København, sid. 54–73.

Scherer, S., Ernst, A., C. Ting-Wei & Böger, P. 1984. Rewetting of drought-resistant blue-green alga: Time course of water uptake and reappearance of respiration, photosynthesis, and nitrogen fixation. – Oecologia 62: 419–423.

Schuber, M. & Kluge, M. 1981. In situ studies of Cras-sulacean Acid Metabolism in Sedum acre L. and Sedum mite Gil. – Oecologia 50: 82–87.

Ström, L., Owen, A. G., Godbold, D. L. & Jones, D. L. 2005. Organic acid behaviour in a calcareous soil – implications for rhizosphere nutrient cycling. – Soil Biol. Biochem. 37: 2046–2054.

Tyler, G. 1971. Hydrology and salinity of Baltic sea-shore meadows. Studies in the ecology of Baltic sea-shore meadows. III. – Oikos 22: 1–20.

Vestergaard, P. (red.) 2007. Det åbne land. – I: Sand-Jensen, K. (red.), Naturen i Danmark. Gyldendal, København.

Zhang, L.-F., Yang, H.-M., Cue, S. X. m.fl. 2009. Proteomic analysis of plasma membranes of cyano-bacterium Synechocystis sp. Strain PCC 6803 in response to high pH stress. – J. Proteomic Res. 8: 2892–2902.

ABSTRACTSand-Jensen, K., Baastrup-Spohr, L., Winkel, A., Møller, C. L., Borum, J., Brodersen, K. P., Lindell, T. & Stæhr, P. A. 2010. Ett kalkbrott på Ölands alvar – en stenöken med knivskarpa miljögränser. [Plant distribution patterns and adaptations in a lime-stone quarry on Öland.] – Svensk Bot. Tidskr. 104: 23–31. Uppsala. ISSN 0039-646X.Extreme environmental gradients generate distinct patterns in species distributions dependent on their stress tolerance and competitive capability. This is the situation along the steep land-water gradi-ent in a stone quarry on Öland’s alvar. Availability of water and nutrients varies profoundly between bare limestone slabs with no soil, via temporary to permanent ponds with thick sediments. Only a few, extremely robust species of plants (Sedum album and S. acre), colonial cyanobacteria (Nostoc com-mune) and mosses (Syntrichia ruralis and Racomitrium canescens) survive on the bare limestone slabs experiencing prolonged summer drought and winter flooding. Nostoc commune can survive drought for many months and resume photosynthesis and effi-cient HCO3

- utilization within minutes after water has become available.

120

STENÖKEN

SVENSK BOTANISK TIDSKRIFT 104:1 (2010) 31

Kaj Sand-Jensen är pro-fessor i akvatisk ekologi vid universitetet i Köpen-hamn och förestår den forskargrupp av lektorer (JB, KPB, PAS), dokto-rander och forskarassis-tenter (LBS, AW, CLM) som besökte Öland för att studera skyfallsalgens

ekologi och fysiologi. Torbjörn Lindell, lektor vid Latinskolan i Växjö, deltog också i projektet.Adress: Ferskvandsbiologisk Laboratorium, Hel-singørsgade 51, DK-3400 Hillerød, DanmarkE-post: ksandjensen@bio.ku.dk

121

 

Supplemetary material 

Included is models, graphs, and tables supporting paper II and paper III. 

122

 

123

Supporting information for: Baastrup-Spohr, L, Iversen, L.L., Borum J, and Sand-Jensen K. (2013) Niche specialization and functional traits regulate abundance of Scandinavian charophytes. Fig. S1. Relationships between the logarithm to number of known localities of charophyte species in Sweden (Sw) and Denmark (Dk) and their corresponding Red List category. Red List categories have been translated to a numerical scale appropriate for statistical analysis: 0 = regionally extinct, 1 = critically threatened, 2 = endangered, 3 = vulnerable, 4 = nearly threatened and 6 = not listed. Spearman rank correlation showed highly significant relationships (p< 0,001) while model II regression yielded no significant difference between the slope of the lines for the two countries P= 0.38. National abundance and Red List categories were taken from Baastrup-Spohr et al. (2013), Langangen (2007), and Gärdenfors (2010).

1 2 3 4 5 60

1

2

3DkSw

Red list category

Abu

ndan

celo

g (n

umbe

r of

loca

litie

s)

Table S2. Functional traits and Scandinavian abundance of the 38 charophyte species that have been found in Sweden, Norway, Finland and Denmark. The functional traits values were obtained from Langangen (2007) and supplemented with values from Gabka (2009). The assignment of a given trait value to each species are explained in detail in the material and methods section of the paper. Abundance data is based on Langangen (2007), Gärdenfors (2010), Koistinen (2010), (Sjøtun et al. 2010) and (Baastrup-Spohr et al. 2013).

124

N/A is indicating that for the given species, the trait value has not been evaluated in the literature

Species Scandinavian abundance

Shoot height

Life form Monoecious/ dioecious

Sexual reproduction

Bulbil production

Salinity tolerance

Chara aspera 22 1 2 1 1 1 18 Chara baltica 18 2 2 0 0 1 17 Chara baueri 0 N/A N/A N/A 2 0 N/A Chara braunii 8 1 0 0 2 0 3 Chara canescens 18 0 0 1 2 0 19.5 Chara connivens 6 0 2 1 1 0 8 Chara contraria 14 1 2 0 2 0 8 Chara curta 1 N/A N/A N/A N/A N/A N/A Chara denudata 2 1 N/A 0 1 0 0 Chara filiformis 1 1 1 0 0 0 0 Chara globularis 24 2 2 0 2 0 7 Chara hispida 14 1 2 0 1 0 1 Chara horrida 6 1 N/A 0 0 0 9.5 Chara intermedia (aculeolata)

16 2 1 0 1 0 1

Chara polyacantha

7 2 2 0 2 0 0

Chara rudis 6 0 1 0 1 0 0 Chara strigosa 12 1 1 0 2 0 0 Chara tomentosa 16 2 1 1 0 1 7.5 Chara virgata 22 1 2 0 2 1 6 Chara vulgaris 14 2 2 0 2 0 8 Lamprothamnium papulosum

7 1 0 0 1 1 17

Nitella batrachosperma

10 0 0 0 2 0 2.25

Nitella capillaris 4 0 0 1 1 0 0 Nitella flexilis 22 2 2 0 1 0 3.5 Nitella gracilis 10 0 2 0 1 0 0.22 Nitella hyalina 3 0 0 0 2 0 2.8 Nitella mucronata

9 1 2 0 0 0 0.47

Nitella opaca 20 2 1 1 1 0 3 Nitella syncapa 2 1 0 1 1 0 0.47 Nitella tenuissima 0 0 0 0 2 0 0 Nitella translucens

7 1 2 0 0 0 0

Nitella wahlbergiana

10 0 0 0 1 0 3.5

Nitellopsis obtusa 10 2 2 1 0 1 8 Tolypella canadensis

12 1 0 0 0 0 0

Tolypella glomerata

5 0 0 0 1 0 0

Tolypella intricata

2 1 0 0 1 0 0

Tolypella nidifica 18 0 0 0 2 0 30 Tolypella normaniana

2 0 0 0 2 0 N/A

125

References Baastrup-Spohr, L. et al. 2013. Rare and threatened charophytes (in Danish). - Urt: in review. Gabka, M. 2009. Charophytes of the Wielkopolska region (NW Poland): distrbution, taxonomy and autecology. - Bogucki Wydawnictwo Naukowe. Gärdenfors, U. (ed.). 2010. The 2010 red list of Swedish species. - Artsdatabanken, SLU. Koistinen, M. 2010. Stoneworts, Characeae. - In: Rassi, P., Hyvärinen, E., Juslén, A. and Mannerkoski, I. (eds.), The 2010 Red List of Finish Species. Ympäristöministeriö & Suomen ympäristökeskus. Langangen, A. 2007. Charophytes of the Nordic countries. - Saeculum ANS. Sjøtun, K. et al. 2010. Cyanophyta, Rhodophyta, Chlorophyta, Ochrophyta. - In: Kålås, J. A., Viken, Å., Henriksen, S. and Skjelseth, S. (eds.), The 2010 Norwegian Red List for Species. Norwegian Biodiversity Information Centre.  

126

Supplementary material: Community assembly Öland Water level modeling In two of the focal ponds the water level fluctuations were modeled using the known water level trace of pond 1 (one measurement per 30 min from 8th of April to 23rd of May) and the manual recordings of water at four dates during the investigated period in the two ponds of unknown trace. Water level was thus modeled for three periods (one between each of the 4 manual recordings). The water level in pond 1 was normalized to the range of values measured during the manual depth recordings using the following equation:

Where NW is the normalized water level, a is the observed depth value while Depthstart cont and Depthend cont is in the start and the end of the period respectively. The modeled water level was calculated as:

Where MW is the modelled water level and Depthstart manual and Depthend manual is the water manually measures at the beginning and end of the each period. For all the focal ponds, there were no continuous measurements occurring within the last 5 days of the investigation period. This period was sunny and without rain and the water level in this period was thus calculated at the linear interpolation between measurements the 23rd of May and the 28th of May. Every vegetation plot (local community) could be directly influenced by the water level of one of the focal ponds due to the very even morphology of the landscape. By leveling each plot according to a focal pond we could construct the water level history of each local community. We condensed the continuous water level of each local community to at mean daily water value(cm), where positive values indicated depth under the water surface and negative values indicated height above the water line. From this data set we constructed several indices that could be involved in determining plant abundances (table 1).

127

Table S1.The abiotic parameters of the local communities were condensed using a principal components analysis. A) the amount of variation explained by the first two principal component axis. B) The eigen vectors of the individual parameters included in the analysis, showing that most variables except bare soil had substantial loadings on the PC1. A PC Eigenvalue % variation accounted

for 1 8.25 58.9 2 2.29 16.4 B Eigen vectors Variable PC1 PC2 days.water 0.301 -0.254 days.cm 0.323 -0.063 wdepth.kum 0.333 -0.128 drought.days -0.257 0.275 longest.wcov 0.309 -0.241 sediment.d 0.240 0.376 SD.sediment.d 0.246 0.352 water.d.may2010 0.317 0.017 veg.h 0.273 0.085 SD.veg.h 0.288 0.069 bryo -0.296 0.171 pebble -0.145 -0.367 bare -0.031 -0.555 Nostoc -0.217 -0.180 Table S2. Plant traits measured on specimens sampled over as much as possible of the environmental gradients in the study area. For all species, the mean value, standard deviation and the number of replicates (n) are given. Plant height was obtained from the LEDA database and values are not shown here.

128

Plant species SLA (cm2 g-1) WLOD (%) LA (cm2) LDMC (mg g-1) Root porosity (%) Agrostis gigantea 152.9 ± 53.3 (34) 35.13 ± 3.34 (3) 0.73 ± 0.24 (34) 327 ± 81.6 (34) 7.6 ± 2.2 (3) Alisma lanceolatum 278 ± * (7) 25.57 ± 5.46 (3) 21.84 ± 3.26 (7) 147.4 ± 40.9 (3) 24.96 ± 4.21 (3) Alisma plantago-aquatile 275.9 ± 33.5 (10) 20.98 ± 3.27 (3) 7.89 ± 2.69 (10) 194.2 ± 37.2 (10) 26.29 ± 2.2 (3) alopecurus geniculatus 253.3 ± 133 (60) 24.26 ± 13.65 (3) 0.75 ± 0.48 (60) 242.1 ± 64.7 (60) 34.33 ± 7 (3) Bromus hordaceus 302.5 ± 41.9 (10) 31.61 ± 4.9 (3) 0.47 ± 0.16 (10) 216.1 ± 22.8 (10) 9.7 6± *** (1) Carex nigra 211.2 ± 31.9 (10) 47.59 ± 6.39 (3) 1.83 ± 0.81 (10) 320.6 ± 21.8 (10) 35.05 ± 9.47 (3) Carex panicea 170.2 ± 29.6 (29) 28.66 ± 4.95 (3) 1.57 ± 0.75 (29) 351.4 ± 60.4 (40) 26.3 ± 1.82 (3) Carex viridula 198.5 ± 55.7 (10) 31.07 ± 6.29 (3) 0.4 ± 0.1 (10) 331.1 ± 62.1 (10) 13.86 ± 1.05 (2) Eleocharis palustris 209.7 ± 117.7 (40) 24.09 ± 6.29 (3) 2.44 ± 1.77 (40) 161.7 ± 31.9 (50) 25.6 ± 7.46 (3) Eleocharis uniglomis 182.7 ± 101.7 (17) 30.81 ± 2.46(3) 2.29 ± 0.89 (17) 171.1 ± 26.2 (17) 27.33 ± 5.41 (3) Equisetum arvense 155.3 ± 20.3 (20) 40.91 ± 3.71 (3) 0.19 ± 0.04 (20) 182.2 ± 15.1 (20) 16.48 ± *** (1) Festuca ovina 143.7 ± 14.6 (10) 52.99 ± 1.09 (3) 0.46 ±0.08 (10) 409.6 ± 51.8 (10) 2.71 ± 3.37 (3) Galium palustre 362.3 ± 78 (17) 61.32 ± 5.39 (3) 0.15 ± 0.07 (17) 186 ± 19.8 (17) 0 ± 2.38 (3) Juncus articulatus 247.4 ± 123 (100) 30.73 ± 3.11 (3)

0.61 ± 0.28 (100)

179.3 ± 45.8 (100) 32.11 ± 3.75 (3)

Juncus Compressus 165.3 ± 71.8 (30) 36.18 ± 15.54 (3) 0.43 ± 0.17 (30) 297.8 ± 44.6 (30) 20.8 ± 3.58 (3) Leontodon autunalis 217.3 ± 60.5 (20) 31.5 ± 1.31 (3) 1.58 ± 0.91 (20) 182.3 ± 33.6 (20) 25.24 ± 2.21 (3) Myriophyllum spicatum 503.6 ± * (10) 76.94 ± 3.71 (3) 0.18 ± 0.06 (10) 132.7 ± 10.5 (10) 26.03 ± 2.61 (3) Plantago lancealata 228.4 ± 69.9 (10) 50.1 ± 8.63 (3) 1.91 ± 0.56 (10) 193.3 ± 95 (10) 0 ± 0.89 (3) Poa anua 243.8 ± 85.8 (20) 47.58 ± 10.2 (3) 0.2 ± 0.07 (20) 278.1 ± 49 (20) 13.58 ± 6.78 (3) Poa compressa 222.8 ± 53.5 (20) 42.16 ± 8.69 (3) 0.49 ± 0.18 (20) 294.7 ± 39.8 (20) 22.2 ± 0.7 (2) Potamogeton natans 287.2 ± 56.1 (15) 67.77 ± 8.81(3) 7.94 ± 3.66 (15) 208.8 ± 22 (19) 18.66 ± 1.91 (3) Potentilla reptans 196.7 ± 36.1 (10) 52.77 ± 6.36 (3) 1.16 ± 0.49 (10) 297.8 ± 24.5 (10) 5.19 ± 1.84 (3) Prunelle vulgaris 229.2 ± 53.3 (20) 33.84 ± 5.55 (3) 0.83 ± 0.41 (20) 212.7 ± 26.6 (20) 10.05 ± 7.33 (3) Ranunculus aquatilis 352.9 ± * (41) 56.12 ± 5.86 (3) 0.08 ± 0.04 (41) 136.2 ± 14.2 (10) 24.52 ± 2.73 (3) Ranunculus flamula 383.3 ± 256.8 (8) 29.31 ± 26.38 (3) 0.28 ± 0.17 (9) 129.1 ± 40.7 (8) 13.81 ± 5.69 (3) Schoenoplectus tabernae-montani 87.8 ± * (4) 13.58 ± 3.04 (3) 7.46 ± ** (16) 164.8 ± 12.7 (6) 31.09 ± 1.73 (3) Sedum acre 160.4 ± 31 (10) 2.85 ± 4.98 (3) 0.09 ± 0.02 (10) 97.7 ± 27.5 (10) 4.89 ± 5.47 (3) Sedum album 112.3 ± 54.9 (50) 6.28 ± 3.08 (3) 0.15 ± 0.08 (50) 94.8 ± 101.5(50) -0.93 ± 9.15 (3) Thymus vulgaris 147.1 ± 20.6(10) 40.29 ± 4.94 (3) 0.09 ± 0.02 (10) 324 ± 22.1 (10) 5.02 ± 6.37 (3)

Typha latifolia 121.9 ± 22.2 (7) 13.47 ± 1.81 (3) 93.92 ± 33.92 (7) 139.7 ± 18.4 (8) 31.08 ± 3.75 (3)

Veronica scutelata 607.3 ± 142.6 (10) 50.52 ± 2.05 (3) 0.75 ± 0.12 (10) 105.8 ± 20.1 (10) 23.97 ± 0.91 (2) Zannichelia palustris 779 ± * (29) 96.93 ± 0.36 (3) 0.03 ± 0.01 (29) 170 ± 107 (5) 27.62 ± 1.35(2) *value based on the stated n-value but because of small leaf size troublesome morphology they were bulk-weighted not enabling calculation of SD. **Estimated on the basis of area length relationship based 4 leaf parts and plant height of 16 plants. ***Only enough root material for one replicate could be sampled because of low biomass of roots. Roots were sampled from more than three individuals.

129

Table S3. Trait clustering index from each trait in each of the 21 investigated local communities. The significance is based on the percentage of null models yielding a higher range in traits than observed. Observed trait ranges smaller than 95% of the null models produced are regarded as significant.

Plot SLA LDMC Root porosity WLOD Height LA 1 0.88 0.82 0.12 2.12 ** -0.20 1.77 * 2 1.31 -0.11 -0.28 0.69 1.15 0.52 3 1.06 0.70 1.83 1.76 * 2.30 ** 0.03 4 0.65 1.57 0.68 -0.19 -0.99 0.23 5 1.70 -1.22 0.17 1.13 1.76 * 2.95 ** 6 -0.65 1.47 1.45 -1.13 -2.44 -1.67 7 -0.12 1.57 2.13 * -0.59 -1.96 -0.84 8 1.95 ** 0.31 0.50 2.25 ** 1.04 2.46 ** 9 -0.04 0.07 -0.04 0.95 1.53 * 2.81 **

10 0.98 -1.28 1.47 -0.24 1.40 1.53 * 11 1.04 -0.15 0.35 1.80 * 1.41 -0.31 12 -1.65 2.35 * 3.44 ** -1.63 -0.38 -1.61 13 1.21 -1.85 -1.21 0.45 1.69 * 1.63 * 14 0.93 0.29 -1.35 0.76 1.66 * -0.29 15 0.47 1.77 2.32 ** 1.44 * 0.56 0.66 16 0.10 1.06 0.77 -1.06 -2.90 0.25 17 -0.33 1.27 3.00 * 0.82 -0.62 -0.09 18 -0.53 0.94 2.36 * 0.71 -0.93 0.36 19 1.55 * -0.38 -1.05 1.13 1.54 1.72 * 20 0.53 -0.55 -0.80 -0.34 1.59 * 0.35 21 0.43 0.50 1.08 0.04 -1.66 -0.39

* ≥ 95 % of null models with a larger range than the observed. ** 100 % of the null models with a larger range than the observed.

130

 

Fig. S5. λ-values of the analyzed traits along the environmental gradient. The λ-values yielded from the maxent models can be interpreted as indicators of trait importance in community assembly. p-values indicate the significance of the correlation between PC1 and the λ-value of a given trait

-4

-2

0

2

4

LD

MC

-15

-10

-5

0

5

W

LOD

-4

-2

0

2

4

6

H

eigh

t

-10

-5

0

5

10

S

LA

-6 -4 -2 0 2 4-6

-4

-2

0

2

4

PC1

Le

af a

rea

-6 -4 -2 0 2 4-4

-2

0

2

4

PC1

P

oros

ity

131

 

131