Arctic plant ecology: From tundra to polar desert in Svalbard - Unis

Arctic plant ecology:
From tundra to polar desert in Svalbard
AB­326 report 2007
Course leaders: Inger Greve Alsos, Christian Körner and David Murray
2007­1
ISBN 978­82­481­0009­6

Content
Arctic plant ecology: From tundra to polar desert in Svalbard – summary
Inger Greve Alsos, Christian Körner and David Murray
1. Biodiversity in the high Arctic: species richness at selected sites in Svalbard,
78­80°N
Henrik Antonsson, Marte Holten Jørgensen and Ane Christensen Tange
2. Life in the Arctic – a struggle for survival?
Simone Lang, Merete Wiken Dees and Kathrin Bockmühl
3. Recruitment along retreating glaciers
Unni Vik, Ingelinn Aarnes and Eike Müller
4. High nutrient levels can compensate for the growth­limiting effect of low
temperatures
Elke Morgner and Christian E. Pettersen
5. Reproductive allocation in the high Arctic
Heike Baldeweg and Emma Bengtsson
6. Freezing resistance in high arctic plant species of Svalbard in mid­summer
Christian Körner, Inger Greve Alsos and AB­326 students
2
3
11
33
55
73
83
93

Arctic plant ecology: From tundra to polar desert in
Svalbard – summary
Inger Greve Alsos 1 , Christian Körner 2 and David Murray 3
1 UNIS – The University Centre in Svalbard, P.O.Box 156, NO­9171 Longyearbyen, Norway, 2 Institute
of Botany, University of Basel, Schönbeinstrasse 6, CH­4056 Basel, Switzerland, 3 University of Alaska
Museum of the North, Fairbanks, Alaska, 99775­6960 USA
Introduction
Within the Arctic, there is a strong climatic gradient from the southern arctic tundra
zone forest ecotone to the polar desert zone found to the very edge of the arctic
landmass (Walker et al. 2005). Three of these zones are present on Svalbard, the mid
arctic tundra zone (zone C), the northern arctic tundra zone (zone B) and the polar
desert zone (zone A, Elvebakk 2005). The mean air temperature during the warmest
month in these three zones in Svalbard are 4­6 ºC , 2.5­4 ºC and 1­2.5 ºC, respectively
(Elvebakk 2005). Because temperature is a major factor determining plant diversity in
the Arctic (Rannie 1986, Walker 1995), global climate warming is expected to have
an impact on arctic flora and vegetation. During our course, the focus was thus on
changes in plant diversity, plant distribution, and plant traits along the climatic
gradient from bioclimatic zone C to A.
Factors other than temperature have strong impacts on plant life and diversity, e.g.
soil moisture (Raup 1969, Gould & Walker 1999), soil acidity (Gould & Walker
1999, Gough et al. 2000, Karlsen & Elvebakk 2003), nutrient input (Körner 2003),
plant productivity (Fox 1992, Jonasson 1992, van der Welle et al. 2003), age of the
terrain (Matthews 1992, Moreau 2005) and natural disturbance (Raup 1969). Our site
selection covered a suite of such co­variables.
Arctic plant communities are expected to undergo species turnover due to climate
change (Walker et al.2006, Kullman 2000, Alsos et al. 2007). This could cause local
loss or increase in plant diversity. Although more than 100 hypotheses have been
raised to address factors determining biodiversity (Grime 1973, Palmer 1994),
variance in species diversity is often poorly explained by predictor variables (Grace
1999). Understanding this issue is importance to predicting short and long­term
effects of climate change on arctic tundra. In project 1, we investigated how soil
temperature, soil moisture, soil pH, exposure, slope angle, bioclimatic zones, and
productivity were related to vascular plant diversity in Svalbard.
Species may have different amplitudes of tolerance to environmental factors (Raup
1969, Karlsen & Elvebakk 2003). Some have wide amplitudes, are broadly tolerant,
whereas others have more narrow amplitude. The question is whether widespread
species have a wide ecological amplitude or could they have a narrow tolerance for
specific habitats, but these habitats are widespread? This is investigated in project 2.
Special emphasis was on Euphrasia wettsteinii, a thermophilic annual which seems to
have spread locally in Svalbard during recent years, a period with warmer than
average temperatures.
3

It is expected that northward expansion of southern species dominating in the Low
Arctic, could cause problems to high arctic species with low competitive abilities.
However, retreating glaciers may provide new habitats in which the high arctic
species may colonise and persist during periods of warmer climate . The ability to
colonise newly deglaciated terrain varies among species (Matthews 1992). In project
3 we studied glacial succession patterns in the three bioclimatic zones to identify high
arctic species that are not represented in up to 100 years old successional stages and
therefore might be at risk of extinction in a warmer climate.
Low temperatures during a short growing season reduce soil nutrient availability.
Since low temperature and low nutrient availability may interact, it is difficult to
distinguish the impact of each of these variables (Körner 2003). Counter­intuitively,
leaves from plants from extremely cold climates had been found to be richer in
macronutients (N, P) than plants from warmer environments, which had been
explained as 'luxurious consumption' (Körner 1989). Locally, the input of nutrient
can be high in Svalbard due to birds or mammals. Especially high levels of nutrients
are found under bird cliffs. The aim of study 4 was to test whether temperature per se
controls plant growth or whether growth is constrained by temperature­induced
nutrient shortage.
In cold climates, plants are generally smaller than in warm climates, and as life
conditions become more adverse miniaturization of plants progresses. Are the
vegetative and reproductive structures similarly affected (isometric allocation), or do
plants give priority to one of the two (asymmetric allocation), which would hint at
selective pressure (evolutionary adaptation)? In alpine environments it has been found
that the allocation of resources shifts towards the reproductive structures as altitudes
increases (Fabbro & Körner 2003). In study 5, we explore this for the first time in the
Arctic.
Climate change is not only expected to result in warmer conditions in the arctic, but
also more extreme weather. This may even include more frequent or more intense
freezing events during the growing period due to changed atmospheric circulation,
despite overall (mean) warming. Freezing resistance is the first and most fundamental
environmental filter plant species have to pass to establish sustainable growth in the
Arctic. However, even arctic plants are susceptible to freezing when they are active,
i.e. during the growing season. Tissues of arctic and alpine plants may tolerate almost
any temperature (some even liquid nitrogen) when they are dormant, but might be
killed by temperatures between ­2 and ­9 °C, when actively growing and flowering.
Hence, it is well established that extreme summer events are the critical issue to look
at (for a review see Körner 2003). During the course, we tested the freezing resistance
of 12 species.
Material and methods
Thirteen students from Norway, Sweden and Germany, studying at nine different
institutions, participated in the course. The students worked in groups of two or three.
All students participated in data collection for all projects, whereas each group was
responsible for study design, analysing and writing about their study. Data were
collected during a seven days cruise with M/S Stockholm. We visited 10 different
4

localities in western and northern Svalbard. We visited one bird cliff and one glacier
foreland in each bioclimatic zone as well as sites with vegetation typical for each
zone.
Two to six sites were visited within each of the ten localities, and a total of 88 plots
were analysed among sites. The plots were randomly chosen for unbiased application
of plot sampling. The point intercept method was used to record the vascular species
within each plot (Bråthen & Hagberg 2004). At each site, 50 cm x 50 cm quadrats
with 25 points of equal intercepts of 10 cm were used. In most cases three replicates
were analysed per sites but sometimes two or four replicates were recorded depending
on time and heterogeneity of the vegetation. Additional species in the plot were also
recorded. The percentage cover of vascular plants, bryophytes, lichens, bare
ground/rocks, and cryptogram crust was estimated for each plot. The following
environmental factors were recorded at each plot: soil temperature, soil moisture, soil
pH, exposure, slope angle, bioclimatic zones, and productivity. Species list for each
site were compiled. These data were shared among project 1 (all sites except the
glacier forelands), 2 (all sites) and 3 (only glacier foreland sites).
To explore whether high nutrient input could compensate for low temperatures, and to
find out how plants allocate resources into growth and reproduction, biometric plant
traits were collected for as many species as possible. Data were collected from 18
species below bird cliffs and in areas with no additional nutrients input. For the
reproduction project, data were collected for a total of 37 species: 17 species in
bioclimatic zone A, 19 species in bioclimatic zone B and 23 species in bioclimatic
zone C. Aboveground plant height, longest leaf length (without petiole) or width (if
the width was larger than the length), and largest blossom diameter (or length of the
whole inflorescence for graminoids) were measured. At least three samples of each
herbaceous species were harvested. In addition, 18 samples of aboveground biomass
was harvested in representative full cover 10 x 10 cm plots, dried and weighed.
For the cold resistance test, at least five mature leaves and five non­senescent, fully
open inflorescences of 12 species were collected for each of the following treatments:
a cool room at +5 °C (control), a freezer set at ­18 °C (guaranteed freezing damage)
and a freezer set to the expected critical range of ­7 °C. Damage was visually
inspected.
The students had ten days for analysing all data and writing their reports. Some
modifications were made on the reports after the course.
Results and discussion
Species diversity, on a landscape as well as at a more local scale, tended to decrease
from bioclimatic zone C to A (Antonsson, Jørgensen and Tange). This can be
explained by a decrease in vegetation cover from bioclimatic zones C to A, as
supported by ANOVA analyses showing significant relations between species density
and the proportion of bare ground to plant cover. In addition, fewer plants are adapted
to the harsh conditions present there (Körner 2003, PAF 2007). This decrease, though
not significant, agrees with previous studies from Svalbard (Marteinsdòttir & Arnesen
2006) and the Arctic (Young 1971, Walker 1995). The relation between species
diversity and the other environmental factors was less clear.
5

Our results suggest that: species diversity, on a landscape as well as local scale,
decreases in relation to bioclimatic zones (from C to A); whereas the number of
species in different vegetation types remains fairly constant throughout the Arctic.
Data on ecological amplitude were collected for 67 vascular plant species (Lang, Dees
and Bockmühl). The most important factors determining distribution of plant species
in Svalbard were pH and bioclimatic zones. There was no species with a pH
amplitude ranging from low to high pH tolerance in bioclimatic zone C, but one and
three in bioclimatic zone B and A, respectively. The ecological amplitude of the
species found in Svalbard were within the range recorded in Greenland (Böcher 1963,
Gelting 1934, Raup 1965, Raup 1969, Sørensen 1933, Karlsen & Elvebakk. 2003). A
minority of species was found to be widely spread and covering a wide amplitudes
regarding pH. The majority of the species studied possessed narrow to medium wide
amplitudes regarding pH, yet they were widespread on Svalbard. No significant
correlation was found between the number of localities where the species was found
and the potential distribution area based on substrate type. Thus, the hypothesis that
widespread species possess a wide ecological amplitude was rejected.
On the glacier forelands in bioclimatic zone C and B, the flora differed on moraines of
different ages (Vik, Aarnes and Müller). Thus, these glacier forelands follow the
succession pattern of species substitution with time as observed in other studies
(Moreau 2005, Hodkinson & Coulson 2003, Jones & Henry 2003). In bioclimatic
zone A however, the vascular plant flora of the moraines were very similar
irrespective of moraines’ age and no clear succession pattern could be found. Species
richness increases in bioclimatic zone C and B with time whilst in bioclimatic zone A
it was comparatively constant. The vegetation in bioclimatic zone A is scattered and
often consists of only single, individual plants. Thus, these marginal populations of
vascular plants rarely develop beyond initial succession phases (Svoboda & Henry
1987). To our knowledge, no other glacial forelands in bioclimatic zone A have been
studied, but we expect that lack of clear, successional stages is a general feature of
glaciers forelands in such extreme environments. The high arctic species Minuartia
rossii, Festuca hyperborea and Poa abbreviata were not found on the glacier
forelands. More glacier forelands should be investigated to see if these habitats could
provide a refugia, at least temporary, for these species in a warmer climate.
Measurements of 778 individuals of 19 different vascular plant species showed a
decrease in plant height, leaf size, and biomass from bioclimatic zone C to A
(Morgner and Pettersen). There was a positive correlation between plant height, leaf
size, and biomass with temperature, as well as nutrient availability. However, there
was also a significant interaction between temperature and nutrient availability in their
effect on plant growth. Increased nutrient levels seemed to compensate for the growth
limiting effect of temperature in bioclimatic zone B and C, but not so in zone A. Thus,
this study provides support to the hypothesis that high nutrient availability can
compensate for the growth limiting effect of low temperatures, but at the thermal limit
for plant growth, temperature itself sets the limit rather than nutrients. Further support
for this hypothesis was evident from the biomass samples. The biomass per unit land
area of full cover plots declined slightly from bioclimatic zone C to A, and bird cliffs
showed much higher biomass than nearby reference sites. The relative fertilizer effect
6

seamed to decrease from zone C to A, but the low number of replicates did not allow
a statistical test of this.
As plant size decreases from bioclimatic zone C to A, the relative investment in
reproductive structure increased in forbs (Baldeweg and Bengtsson). Graminoids on
the other hand, change isometrically along the same gradient. This is probably due to
different pollination systems. Insect pollinated species have to attract pollinators in an
increasingly barren landscape with overproportionally showy flowers. Pollinator­
driven pollination thus appears to select inflorescence size differently from wind­
driven pollination.
The cold resistance test showed that the 12 species tested exhibited a broad damage
spectrum from zero to 100% (Körner and Alsos). We concluded that during the
growing season, freezing temperatures of ­7 °C would damage a significant fraction
of the arctic flora. Given that ­10 °C has been recorded for air temperature in
bioclimatic zone C in June, this means that earlier spring growth in the course of
global warming at an otherwise unchanged likelihood of the occurrence of such an
extreme events would lead to massive losses of tissue and above ground productivity.
Our survey confirms knowledge from the Alps for the critical range of temperature to
be explored in a more detailed assessment The study also indicates that it matters a
lot which species are examined. Overall, the data appear to match the summer
freezing tolerance known for alpine plants of the temperate zone (Körner 2003) and
do not indicate a greater frost hardiness for these high arctic plants as had been
suggested from experiments with Saxifraga oppositifolia under controlled growth
conditions (Robberecht and Junttila 1992).
References
Alsos, I.G., Eidesen, P.B., Ehrich, D., Skrede, I., Westergaard, K., Jacobsen, G.H.,
Landvik, J.Y., Taberlet, P., Brochmann, C. 2007: Frequent Long­Distance Plant
Colonization in the Changing Arctic. Science 316: 1606 – 1609
Bråthen, K. A. & Hagberg, O. 2004: More efficient estimation of plant biomass.
Journal of Vegetation Science, 15: 653­660
Böcher, T. W. 1963: Phytogeography of middle west Greenland. Meddelelser om
Grønland, 148
Elvebakk, A. 2005: A vegetation map of Svalbard on the scale 1 : 3.5 mill.
Phytocoenologia 35: 951­967
Fabbro T. & Körner C. 2004: Altitudinal differences in flower traits and reproductive
allocation. Flora 199: 70­81
Fox, J.F. 1992: Responses of diversity and growth form dominance to fertility in
Alaskan tundra fellfield communities. Arctic and Alpine research 24:233­237
Gelting, P. 1934: Studies on the vascular plants of east Greenland between Franz
Joseph Fjord and Dove Bay (Lat. 73° 15’ – 76° 20’ N.). Meddelelser om Grønland,
101
7

Gough, L., Shaver, G.R., Carroll, J., Royer, D.L. & Laundre, J.A. 2000: Vascular
plant species richness in Alaskan arctic tundra: the importance of soil pH. Journal of
Ecology 88: 54­66
Gould, W.A. & Walker, M.D. 1999: Plant communities and landscape diversity along
a Canadian Arctic river. Journal of Vegetation Science 10: 537­548
Grace, J. B. 1999: The factors controlling species density in herbaceous plant
communities: an assessment. Perspectives in Plant Ecology, Evolution and
Systematics 2: 1­28
Grime, J.P. 1973: Control of species density in herbaceous vegetation. J. Environ.
Manage. 1: 151–167
Hodkinson, I. D., Coulson, S. J. & Webb, N. R. 2003: Community assembly along
proglacial chronosequences in the high Arcic: vegetation and soil development in
north­west Svalbard. Journal of Ecology 91, 651­663
Jonasson, S. 1992: Plant­responses to fertilization and species removal in tundra
related to ommunity structure and clonality. Oikos 63: 420­429
Jones. G. & Henry, G. H. R. 2003: Primary plant succession on recently deglaciated
terrain in the Canadian High Arctic. Journal of Biogeography 30, 277­296
Karlsen, S. R. & A. Elvebakk 2003: A method using indicator plants to map local
climatic variation in the Kangerlussuaq/Scoresby Sund area, East Greenland.
Journal of Biogeography 30: 1469­1491
Kullman, L. 2000. Tree­limit rise and recent warming: a geoecological case study from the
Swedish Scandes. Norsk. geografisk Tidsskrift 54: 49­59
Körner C. 1989. The nutritional status of plants from high altitudes. A worldwide
comparison. Oecologia: 81: 379­391
Körner, C. 2003: Alpine plant Life. Springer, Berlin
Marteinsdóttir, B. & Arnesen, G. 2006: Species richness of vascular plants and
cryptogames in the High Arctic: no universal pattern observed in relation to primary
productivity. Pp. 11­22 in I.S. Jónsdóttir: Exploring plant­ecological patterns at different
spatial scales in Svalbard. UNIS publication series 2006­1, www.unis.no
Matthews, J. A. 1992: The ecology of recently deglaciated terrain. A geoecological
approach to glacier forelands and primary succession. Cambridge, Cambridge
University Press
Moreau, M., Laffly, D., Joly, D. & Brossard, T. 2005: Analysis of plant colonization
on an arctic moraine since the end of the little Ice Age using remotely sensed data
and a Bayesian approach. Remote Sensing of Environment 99, 244­253
PAF. 2007: Checklist of the Panarctic Flora (PAF). Draft version. I: Elven, R.,
Murray, D. F., Razzhivin, V. & Yurtsev, B. A. (red.). Oslo, Univ. of Oslo
Palmer, M.W. 1994: Variation in species richness: towards a unification of
hypotheses. Folia Geobotanica & Phytotaxonomica 29: 511–530
Rannie, W. F. 1986: Summer air temperature and number of vascular species in arctic
Canada. Arctic 39: 133­137
8

Raup, H. M. 1965: The flowering plants and ferns of the Mesters Vig district,
northeast Greenland. Meddelelser om Grønland, 166
Raup, H. M. 1969: The relation of the vascular flora to some factors of site in the
Mester Vig district, northeast Greenland. Meddelelser om Grønland, 176: 1­80
Robberecht, R. & Junttila, O. 1992: The Freezing Response of an Arctic Cushion
Plant, Saxifraga caespitosa L.: Acclimation, Freezing Tolerance and Ice Nucleation.
Ann Bot 70:129­135
Svoboda, J., Henry, G. H. R. 1987: Succession in marginal arctic environments.
Arctic and Alpine Research 19, No 4, 373­384
Sørensen, T. 1933: The vascular plants of East Greenland from 71° 00’ to 73° 30’ N.
Meddelelser om Grønland. 101
Walker, M. D. 1995: Patterns and causes of Arctic plant community diversity. Arctic
and alpine biodiversity: Pattern, causes and ecosystem consequences. In : F. S.
Chapin III and C. Körner. Berlin, Springer­Verlag: 3­20
Walker, M. D., Wahren, H. C., Hollister R. D., Henry, G. H. R., Ahlquist, L. E.,
Alatalo,
J. M., Bret­Harte, S. M., Calef, M. P., Callaghan, T. V., Carroll, A. B., Epstein, H.
E., Jónsdóttir, I. S., Klein, J. A., Magnússon, B., Molau, U., Oberbauer, S. F., Rewa
S. P., Robinson, C. H., Shaver, G. R., Suding, K. N., Thompson, , C. C., Tolvanen,
A., Totland, Ø., Turner, P. L., Tweedie, C. E., Webber, P. J. & Wookey, P. A. 2006:
Plant community responses to experimental warming across the tundra biome.
Proceedings of the National Academy of Sciences of the Unites States of America
103, 1342­1346
van der Welle, M.E.W., Vermeulen, P.J., Shaver, G.R. & Berendse, F. 2003: Factors
determining plant species richness in Alaskan arctic tundra. Journal of Vegetation
Science 14: 711­720
Young S.B. 1971: The vascular flora of St. Lawrence Island with special reference to
floristic zonation in the arctic regions. Contrib Gray Herb 201:11­115
9

Biodiversity in the High Arctic: species richness at
selected sites in Svalbard, 78­80°N
Henrik Antonsson 1 , Marte Holten Jørgensen 2 , and Ane Christensen Tange 3
1 Department of Plant and Environmental Sciences, Göteborg University. Box 460 S­405 30 Göteborg,
Sweden. E­mail: henrik.antonsson@dpes.gu.se. 2 Program for Molecular Ecology and Biosystematics,
Department of Biology, University of Oslo, P.O. Box 1066 Blindern, NO­0316 Oslo, Norway.
3 Department of Biology, Norwegian University of Science and Technology, NTNU NO­7491
Trondheim, Norway.
Abstract
Biodiversity is today greatly endangered by human activities, particularly in the
vulnerable Arctic. We studied 15 plant communities in the arctic archipelago of
Svalbard (78­80°N) distributed in three different bioclimatic zones in search for
differences in species richness, and causes for these. We recorded biodiversity as point
frame interceptions (0.25 m 2 ; 25 points), species present in frames, and species lists
for each site visited, and compared them with major environmental factors; soil
temperature, soil moisture, soil pH, exposure, slope angle, bioclimatic zones, and
productivity. The resulting dataset was analysed using Kruskal­Wallis tests, Kendall’s
τ correlations, ANOVA and ordination methods (PCA and RDA). Our results suggest
that: species diversity, on a landscape as well as local scale, decreases in relation to
bioclimatic zones (from C to A). Ordination methods implied that species richness is
correlated with bioclimatic zones, soil temperature, production and ground cover. We
conclude that biodiversity in the Arctic is highly scale­dependent and that this model
needs more testing, in order for policymakers to make wise decisions on conservation
of Arctic environments.
Keywords: Arctic, bioclimatic zones, biodiversity, plants, species richness, Svalbard
Introduction
Human activities are reducing the number of species worldwide at a rapid pace with
potentially devastating effects on ecosystem services, and in the long run economical
and social consequences (Hooper et al. 2005). Climate changes induce migration of
species, loss of some species, replacement of some and arrival of invading species into
plant communities. Proof for this pattern has been published from numerous sites
along the tundra and the Arctic, from experimental warning experiments (Walker et
al.2006), as well as real changes in vegetation (Kullman 2000, Truong 2003). This
may cause large­scale changes in community composition, which surely will have a
detrimental effect on arctic and alpine ecosystems. Concerns for loss of biodiversity
are raised throughout the world and investigations on what factors are most important
in determining biodiversity are heavily debated in the scientific community. Increasing
our knowledge on this issue is not merely of academic interest, but of great importance
in order to predict short and long­term effects of climate change on arctic tundra.
The first general model concerning what determines plant diversity was developed by
Grime (1973). His humpback or intermediate disturbance model of plant species
density predicted that species richness is depending on level of stress and production.
Following Grime, a large number of studies on this topic have been published, most of
11

them supporting his theory (e.g., Huston 1979, Tilman 1982, Taylor 1990). In
summary, the majority of theories and hypotheses accentuate intermediate levels of
productivity, disturbance, pH and other abiotic factors as yielding the highest levels of
species richness. The issue is, however, very complex, and more than 100 hypotheses
on variables controlling species richness have been presented throughout the years
(Palmer 1994), and variance in species density is often poorly explained by predictor
variables (Grace 1999).
Although containing a vascular plant flora of about 2000 taxa of vascular plants
(Elven et al. 2007), and at least another 1000 taxa of bryophytes and lichens, the arctic
flora contains fewer species than ecosystems of lower latitudes (Gaston et al. 1998).
Walker (1995) propose that there is a fine­pored filter of the extreme environmental
conditions present in the Arctic, effectively excluding many plant species from
establishment in this region. Studies on variation of species richness along abiotic
gradients in arctic and alpine communities point out temperature (Young 1971, Rannie
1986, Walker 1995), soil moisture (Gould & Walker 1999, Michalet et al. 2002), soil
acidity (Gould & Walker 1999, Gough et al. 2000, Roem & Berendse 2000),
productivity (Fox 1992, Jonasson 1992, Theodose & Bowman 1997, Gough et al.
2002, Grytnes 2002, van der Welle et al. 2003, Constanza et al. 2007) as important
factors for species richness, and that local conditions play a crucial role (e.g., Gough et
al. 2000).
Our aim was to investigate the major factors determining vascular plant diversity in
different communities along a latitudinal gradient in the High Arctic, expecting a
decrease in species richness moving northwards through the bioclimatic zones C, B,
and A (CAVM Team 2003). From our general hypothesis, the following predictions
were outlined: Biodiversity is correlated with bioclimatic zone, community biovolume
(as a substitute for biomass) is correlated with bioclimatic zone and species richness,
and species richness is correlated with soil moisture and soil acidity.
Material and methods
Study sites and data collection
Sites were chosen according to latitude, bioclimatic zone (Elvebakk et al. 1999,
CAVM team 2003, Elvebakk 2005), vegetation type, and microclimatic conditions
along the western and northern coast of Svalbard (Table 1; Figure 1).
In our study we included Florabukta as bioclimatic zone A, even though it is in the
border between zone A and B. For vegetation analysis three different measures were
used for biodiversity; 1.) Species density: noting every vascular plant individual hit
when using a needle and point frames (50x50 cm 2 ; 25 points; see Bråthen & Hagberg
2004), 2.) Species richness: noting all vascular plant species found within the frame
and 3.) Total species list: noting all vascular plant species found in each site, an area of
roughly 300 x 300 m, consisting of the same vegetation type. For each frame the
following abiotic factors were measured; slope angle, exposure, organic soil depth,
soil temperature, soil moisture, and soil pH. Slope angle was measured with an
inclinometer with the precision of 2°; exposure was recorded with a compass, defined
as the direction in which the slope was facing; organic soil depth was measured with a
ruler after digging a soil profile, soil temperature was measured with a thermometer at
approximately 10 cm depth to calibrate for recent and rapid changes. Soil moisture
was estimated in the field by digging the fingers in the soil and determining moisture
12

along a scale where 1 = dry, 2 = mesic, and 3 = wet. For soil pH, soil samples were
taken in the field and analyzed indoors with a pH meter. Cover was estimated in the
field as percentage of the frame, for vascular plants, lichens, bryophytes, cryptogam
crust, and bare ground. In addition, mean canopy height of vascular plants was
estimated (in centimetres) in the frame. As a substitute for biomass, vascular
biovolume was calculated for each site and frame as the canopy height (cm) multiplied
by vascular plant cover (%). It has been shown in a study by Lindblad (2007), that
vascular plant height was a better measure than biomass to explain vascular plant
diversity in a mid­alpine, Sub­Arctic plant community. Similarly, Gough et al. (2000),
found a unimodal­shaped correlation of canopy height and species richness and
density. This suggests that using biovolume is a measure at least as good as biomass.
Lady Franklinfjorden
Biscayarhuken
Florabukta
Reinsdyrflya
Mayerbreen
Fjortende julibreen
Ossian Sarsfjellet
Colesdalen
Kapp Thordsen
Figure 1. Sites included in this study.
At each site, at least three replicate frame analyses were made (Table 1). For some of
the measures (cover and soil moisture), subjective estimations were used. This was
carried out by a large group of people taking turns and the preceding calibration of
methods may have been insufficient to make sure that everybody carried out
measurements in the exact same way. This should be kept in mind when interpreting
the data.
13

Site ID
Table 1. Sites analysed in this study.
Name
Date
1607/1/1 Kapp Thordsen 16.07.2007 8709800, 33X 0511516 C Mesic heath 3
1607/1/2 Kapp Thordsen 16.07.2007 8709858, 33X 0511838 C Calcarious fen 3
1707/1/2­3 Mayerbreen 17.07.2007 8800608, 33X 0440870 B Old moraine area 7
1707/2/1 Fjortende Julibreen 17.07.2007 8786090, 33X 0439992 B Birdcliff 3
1707/2/2 Fjortende Julibreen 17.07.2007 8786090, 33X 0433912 B Dry heath 3
1807/1/1­2 Biscayarhuken 18.07.2007 8864496, 33X 0448709 B Dry heath 5
1807/2/1­2 Reinsdyrflya 18.07.2007 8863020, 33X 0478260 B Mesic heath 6
1907/1/1 Florabukta 19.07.2007 8888054, 33X 0571330 A Birdcliff 3
1907/1/2 Florabukta 19.07.2007 8887920, 33X 0571369 A Mesic heath 3
1907/1/3 Florabukta 19.07.2007 8887344, 33X 0571431 A Polar desert 3
2007/1/3 Lady Franklinfjorden 20.07.2007 8901644, 33X 0587985 A Mesic heath 3
2107/1/3 Ossian Sarsfjellet 21.07.2007 8762902, 33X 0445660 C Mesic heath 3
2107/2/1­2 Ossian Sarsfjellet 21.07.2007 8763388, 33X 0445378 C Birdcliff 4
2207/1/1­5 Colesdalen 22.07.2007 8670162, 33X 0502830 C Mesic heath 18
2207/1/6 Colesdalen 22.07.2007 8670038, 33X 0502916 C Mire 4
14
UTM
coordinates
Bioclimatic
zone
Vegetation
type
Data analysis
Shannon’s diversity index H was calculated using the species density results and the
statistical package PAST (Hammer et al. 2001). The index was used as a measure of
species diversity in all analyses. Kruskal­Wallis one­criterion variance analysis
(Kruskal & Wallis 1952) was run with the same program, grouping the plots according
to bioclimatic zones a priori (zone A=1, B=2, C=3). Both species composition and
number of species was tested, as well as a comparison of the different sites. Kendall’s
τ correlations (e.g., Conover, 1998) were calculated for chosen measures using PAST.
Significance level was reduced with Bonferonni corrections for multiple tests of
independent variables (e.g., Abdi, 2007). The species richness dataset was clustered
using simple matching similarity and unweighted pair group method with arithmetic
mean (UPGMA) in NTSYSpc 2.02h (Rohlf, 1999).
Ordination analyses of results per site were performed using Canoco 4.5 (Biometris –
Plant Research International, Wageningen). Environmental factors were chosen based
on the Kendall’s τ correlations, and all data included were transformed
logarithmically. A detrended correspondence analysis (DCA) was run to test
distribution of the data. Being linear, the species data (Shannon’s H, species density,
species richness, and total species list) were analysed with a principal component
analysis (PCA), and environmental data were added a posteriori. Redundancy analysis
(RDA) was used to find explanatory effects of the environmental factors, and these
were tested with a Markov Chain Monte Carlo (MCMC) permutation test.
Analysis of variance (one way ANOVA) was run using S PLUS 6.2 for Windows
(Insightful Corp.). All data were transformed logarithmically before running the
analysis. All species measures were compared with the abiotic measures included in
the ordination analyses. In the ANOVA, factors were divided into evenly sized groups.
Tukey HSD post hoc analyses were run when ANOVA gave significant results (for
No. of plots

details see Crawley 2002). The PCA, RDA and ANOVA analysis were performed
using site means in order to include the total species list.
Results
Species richness, species density and total species list are summarised in Figure 2, and
complete lists are in Appendix 1. We found 25, 32, and 49 species in total in zones A,
B, and C, respectively. More species were registered using presence/absence in frame
(species richness) as measure, than the point frame hits (species density) in all sites
(Figure 2). The Kruskal­Wallis test gave no significant results when grouping the
species richness per frame according to bioclimatic zones (H=5.039; Hc=5.11;
p=0.0805). Species richness was, however, significantly different among sites
(H=30.48; Hc=30.91; p=0.0066).
Figure 2. Summary of species data. Box plot of species density, species richness and total
species lists grouped according to bioclimatic zones.
Kendall’s τ correlations are given in Table 2. The only biodiversity measure
significantly correlated with any environmental factors, was species density which was
positively correlated with bioclimatic zones, meaning that the number of species
decreases going from zone C to B to A, and negatively correlated with latitude.
The UPGMA analysis clustered the sites more or less according to bioclimatic zones
and vegetation types (not shown). The sites from zone A and B were clustered
together and included also a small cluster with zone C (the heath sites 1607/1/1 and
2107/1/3, and the calcareous fen site 1607/1/2), whereas the remaining sites from zone
C clustered together; the species rich sites in Colesdalen (2207) and the birdcliff site in
Ossian Sarsfjellet.
15

Table 2. Kendall’s τ correlations among selected measures. The lower triangle gives the correlation coefficients, the upper gives the probabilities. Significance is
calculated with Bonferonni corrections for multiple tests (p

The significant results from ANOVA are given in Table 3. The Shannon’s H index as
well as the species density was highest where there was low (0.00­6.00%) and
intermediate (7.00­36.00%) amount of bare ground and in sparse (10.00 cm) organic soil.
Table 3. Significant results from ANOVA analysis.
Bare ground cover Organic soil depth
df F­value p­value df F­value p­value
Shannon’s H 2 7.56 0.008 3 5.36 0.016
Species density 2 6.65 0.011 3 5.65 0.013
The PCA ordination showed correlation between species richness and total species list,
and between Shannon’s H and species density (Fig. 3). 81.6% of the variation in the
data was explained by the first axis and 11.8% by the second axis. There was a
positive correlation between all of the following variables: organic soil depth,
bioclimatic zone, temperature and vascular biovolume. Bare ground was negatively
correlated with the variables mentioned above. pH was not correlated with the other
environmental variables. The sites were not significantly distributed according to the
environmental data.
­0.8 1.0
total species list
species richness
13
org. soil depth
2
bioclimatic zone
14 4
9
15
Shannon's H
species density
temperature 8
1
pH
­1.0 1.0
17
7
3
vasc. biovolume
5
6
12
10
bare ground
Figure 3. PCA analysis of biodiversity and chosen environmental data. Axis 1 spans 81.6% of
the variation and axis 2 spans 11.8%. The sites are marked according to bioclimatic
zones: A – black, B – grey, and C – white.
11

The RDA ordination (Fig. 4) showed overall the same trends as the PCA ordination.
The MCMC run in RDA gave no significant result for the degree of explanation
(p=0.0740) from the environmental data. 26.0% of the variation in the data was
explained by the first axis and 4.4% of the variation in the data was explained by the
second axis in RDA. The sites were somewhat separated according to bioclimatic
zones along the first axis, and bioclimatic zones explained 20% of the variation in the
MCMC test, although not significantly (p=0.0740). All measures of diversity showed a
positive correlation with increasing bioclimatic zone (from A to C), although not
significantly. There was a relationship between bioclimatic zones and the degree of
bare ground and the layer of organic soil. There was less bare ground (3 ± 2 %) and a
deeper layer of organic soil (10.21 ± 6.21 cm) in zone C, than zone A (49 ± 35 %, 2.54
± 2.95 cm, respectively). Zone B was intermediate (12 ± 13 %, 5.00 ± 3.22 cm,
respectively). There was also a trend in increased soil temperature and vascular
biovolume in zone C (6.59 ± 1.68ºC, 1.56 ± 0.58, respectively) compared to zone A
(3.93 ± 0.49ºC, 0.39 ± 0.25, respectively).
­1.0 1.0
11
8
bare ground
10
9
3
5
7
vasc. biovolume
6
13
temperature
12
4
­0.8 1.0
14
18
Shannon's H
species density org. soil depth
bioclimatic zone
15
2
species richness
pH
total species list
Figure 4. RDA analysis of biodiversity and chosen environmental factors. Axis 1 spans 26.0%
of the variation, and axis 2 spans 4.4%. The sites are marked according to bioclimatic
zones: A – black, B – grey, and C – white.
Discussion
Our measures of biodiversity (species richness, species density, Shannon’s H index
and total species list) did not correspond perfectly to the different abiotic variables, but
we feel the following statements are robust: 1) the value of the total species list is
1

different among the three bioclimatic zones; 2) mean species richness per site is not
significantly different among zones; and 3) species density is different among zones.
Important when discussing biodiversity is the scale on which investigation is made.
Comparing biodiversity among sites is only possible when using the same scale. All
three of our measures of biodiversity are records of α­diversity, defined as the
diversity within a stand or community. To make a meaningful comparison between
regions, it is important to operate on the same scale.
The total species list showed an increase in numbers when moving from bioclimatic
zone A to C. This trend, though not significant in our study, is in line with previous
studies from Svalbard (Marteinsdòttir & Arnesen 2006) and the Arctic in general
(Young 1971, Walker 1995). This is a measure of the diversity at landscape level,
including several sites with different vegetation types. Comparing our data with the
database for vascular plants found on Svalbard (Svalbarddatabasen), there is a total of
110 vascular plants recorded from Colesdalen, situated in bioclimatic zone C, while
Gustav V Land (zone A) has a record of 85 species, although the area is significantly
larger. See also Table 4 for comparison with other arctic areas.
On a smaller scale, when looking at mean species richness per site, there is no
significant difference between sites in different regions in our study. Most sites
contain between 17 and 27 species, which is correlating very well with a study from
the Canadian Arctic by Franzen & Molau (1999). They found no trends of changes in
species richness when going from latitude 62° to 78°. In our study, covering a large
gradient, we visited a restricted number of different vegetation types, and hence could
not capture the full plant diversity of every site. In the field the plots were chosen to
cover one particular kind of habitat (vegetation type) in each site.
Species density (mean per site) is increasing when going from zone A to C (Table 2,
Figure 2). This can be explained by more scattered vegetation in bioclimatic zones A
and B compared to C (Table 2), as supported by ANOVA analyses showing significant
relations between species density and bare ground, as well as the fact that fewer plants
are adapted to the harsh conditions present there (Körner 2003).
Our results largely supported our first prediction, that biodiversity is correlated with
bioclimatic zone and highlight the importance of scale when discussing biodiversity,
as shown in the discussion above. From our results, and also when comparing with
Franzen & Molau (1999), it seems likely that most vegetation types in the Arctic
contain roughly the same number of species, but differences among regions remain on
a larger scale (among bioclimatic zones), due to the degree of landscape heterogeneity
and the number of different habitats available for plant species (see Table 4).
The first part of our second prediction, concerning community biovolume and
bioclimatic zone also gained support from our data. We found community biovolume
to decrease with bioclimatic zone, but increased biovolume did not significantly relate
to any measure of biodiversity. This is consistent with the work of van der Welle et al.
(2003), and Gough & al. (2000), which showed no correlation of biodiversity and
productivity in the Arctic. On the other hand, a large number of studies support the
view of a ‘hump back’ relationship of productivity and species richness in alpine and
arctic areas (Grytnes 2002, Theodose & Bowman 1997, Jonasson 1992). Constanza et
19

al. (2007) found that at low temperatures (−2.1 °C annual mean), biodiversity was
negatively correlated with net primary production. In Svalbard the annual mean
temperature range from ­7°C to ­15°C, so our study does not confirm this view. In a
fertilization experiment in boreal­alpine Alaska, Fox (1992) found a change in growth
form and species composition rather than effects on species richness. It is possible that
community productivity, as well as other abiotic variables, may have a larger influence
on species composition, rather than direct effects on species richness. Although our
different measures of biodiversity is not clearly concordant, the observed pattern was
that the dominant species, mainly woody perennials and dwarf shrubs, were the first to
disappear in the areas with harsher environmental conditions (zone B and A), whereas
a fairly large number of species persist, but with very few individuals.
Our prediction concerning pH and species richness was not supported by our data,
although the relationship between pH and species richness has been positively
correlated in numerous studies from different vegetation types, such as arctic tundra
(Gough et al. 2000, Gould & Walker 1999), temperate heath­ and grassland (Roem &
Berendse 2000) as well as Svalbard (Wenche et al. 2002). Our sample sites were,
however, distributed within a relatively narrow range of pH (5.5­8.5), thus the full
effect of pH on species richness could not be studied.
The prediction on soil moisture was not supported in our study. Explanations for this
could be that it was measured on a coarse scale, wherein most estimations of moisture
turned out to be medium. In addition these measures also were heavily influenced by
precipitation in the preceding 24 hours.
Our ordination tests (PCA and RDA) gave no significant results, but indicated which
factors are most important in explaining species richness. Bioclimatic zone and
organic soil depth explain the increased species diversity. Bare ground cover is
negatively correlated to all of our diversity measures. This is also in line with the
results of our ANOVA. As to why we could not get a significant explanation, we
suggest that different factors may be acting with different strength in different zones.
Conclusion
Differences in biodiversity are found when looking at a landscape level and not at plot
(α­) level, among frames at different sites. Hence, what creates high biodiversity in the
Arctic seems to be variation of habitats and vegetation types on the landscape. We
welcome further studies on this explanation model, and conclude that if supported by
more general data, this may be an important finding concerning vulnerable Arctic
ecosystems. If landscape heterogeneity is the major source of Arctic biodiversity,
policymakers need to take this into consideration for future conservation of the Arctic
environment
Acknowledgements
We thank Inger Greve Alsos and David F. Murray for advices and comments through
all the process, and the participants in the course AB326 Arctic Plant Ecology 2007
for fruitful discussions.
References
Abdi, H. 2007: The Bonferonni and Šidák corrections for multiple comparisons. In:
Salkind, N. (Ed), Encyclopedia of Measurement and Statistics. Sage, Thousand Oaks
20

Barrett, P.E. & Teeri, J.A. 1973: Vascular plants of the Truelove Inlet region, Devon
Island. Arctic 26: 58­67
Bay, C. 1992: A phytogeographical study of the vascular plants of northern Greenland
­ north of 74 degrees northern latitude. Meddelelser om Grønland Bioscience 36
Beschel, R.E. 1963: Geobotanical studies on Axel Heiberg Island in 1962. Pp. 1­18 in:
Müller, F. (Ed), Axel Heiberg Preliminary Rept. 1961­1962. McGill University
Billings, W.D. 1992: Phytogeographic and evolutionary potential of the arctic flora in
a changing climate. Pp: 59­89 in: Chapin FS III, Jefferies RL, Reynolds, JF, Shaver
GS, Svoboda J (Eds) Arctic ecosystems in a changing climate: an ecophysiological
perspective. Academic Press, San Diego
Bird, C.D. 1975: The lichen, bryophyte, and vascular plant flora and vegetation of the
Landing Lake area, Prince Patrick Island, arctic Canada. Canadian Journal of Botany
53: 719­744
Brassard, G. & Beschel, R.E. 1968: The vascular flora of Tanquary Fiord, northern
Ellesmere Island, N.W.T. Canadian Field­Naturalist 82: 103­113
Brassard, G. & Langton, R.E. 1970: The flora and vegetation of Van Hauen Pass,
northwestern Ellesmere Island. Canadian Field­Naturalist 84: 357­364
Bruggermann, P.F. & Calder, J.A. 1953: Botanical investigation in northeast
Ellesmere Island, 1951. Canadian Field­Naturalist 67: 157­174
Bråthen, K.A. & Hagberg, O. 2004: more efficient estimation of plant biomass.Journal
of Vegetation Science 15:653­660
CAVM(Circumpolar Arctic Vegetation Map) Team. 2003: Circumpolar Arctic
Vegetation Map. Scale 1:7 500 000. Conservation of Arctic Flora and Fauna (CAFF)
Map no. 1 U.S. Fish and Wildlife service, Anchorage, Alaska
Conover, W.J. 1998: Practical non­parametric statistics. John Wiley and Sons, New
York
Costanza, R., Fisher, B., Mulder, K., Liu, S. & Christopher, T. 2007: Biodiversity and
ecosystem services: A multi­scale empirical study of the relationship between
species richness and net primary production. Ecological Economics 61: 478­491
Crawley, M. J. 2002: Staistical computing: an introduction to data analysis using S­
plus. John Wiley and Sons Ltd., Chichester
Elvebakk, A. 2005: A vegetation map of Svalbard on the scale 1:3.5 mill.
Phytocoenologia 35: 951­967
Elvebakk, A., Elven, R. & Razzhivin, V.Y. 1999: Delimitation, zonal and sectorial
subdivision of the Arctic for the Panarctic Flora Project. Det Norske Vitenskaps­
Akademi. I. Matematisk Naturvitenskapelig Klasse. Skrifter. Ny serie 38: 375­386
Elven, R., Murrey, D.F., Razzhivin, V. & Yurtsev, B.A. 2007: Checklist for the
Panarctic Flora (PAF): Vascular Plants. Draft Version. University of Oslo, Oslo
Fox, J.F. 1992: Responses of diversity and growth form dominance to fertility in
Alaskan tundra fellfield communities. Arctic and Alpine research 24:233­237
21

Franzèn, D. & Molau, U. 1999: Vascular plant diversity patterns in the Canadian
Arctic. Polarforskningssekretariatets årsbok 1999: Stockholm
Gaston, K.J, Blackburn, T.M, & Spicer, J.I. 1998: Rapoport’s rule: time for an
epitaph? Trends in Ecology and Evolution 13: 70­74
Gough, L., Shaver, G.R., Carroll, J., Royer, D.L. & Laundre, J.A. 2000: Vascular plant
species richness in Alaskan arctic tundra: the importance of soil pH. Journal of
Ecology 88: 54­66
Gould, W.A. & Walker, M.D. 1999: Plant communities and landscape diversity along
a Canadian Arctic river. Journal of Vegetation Science 10: 537­548
Grace, J.B. 1999: The factors controlling species density in herbaceous plant
communities: an assessment. Perspectives in Plant Ecology, Evolution and
Systematics 2: 1­28
Grime, J.P. 1973: Control of species density in herbaceous vegetation. Journal of
Environmental Management 1: 151–167
Grytnes, J.A. 2002: Fine­scale vascular plant species richness in different alpine
vegetation types: relationships with biomass and cover. Journal of Vegetation
Science 11:87­92
Hammer, Ø., Harper, D.A.T. & Ryan, P.D. 2001: PAST: Paleontological statistics
software package for education and data analysis. Palaeontologia Electronica 4: 9 pp
Hooper, D.U., Chapin, F.S. III, Ewel, J.J., Hector, A., Inchausti, T., Lavorel, S.,
Lawton, J.H., Lodge, D.M., Lorau, 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: 3­35
Huston, M.A. 1979: A general hypothesis of species diversity. American Naturalist
113: 81–101
Jonasson, S. 1992: Plant responses to fertilization and species removal in tundra
related tocommunity structure and clonality. Oikos 63: 420­429
Kullman, L. 2000: Tree­limit rise and recent warming: a geoecological case study
from the Swedish Scandes. Norsk. geografisk Tidsskrift 54: 49­59
Körner, C. 2003: Alpine plant life : functional plant ecology of high mountain
ecosystems : Springer Verlag, Berlin
Kruskal, W.H. & Wallis, W.A. 1952: Use of ranks in one­criterion variance analysis.
Journal of the American Statistical Association 47: 583­621
Lindblad, K. 2007: Tundra Landscape Ecology – Diversity across scales. PhD thesis, Göteborg
University. pp 48
Marteinsdòttir, B. & Arnesen, G. 2006: Species richness of vascular plants and
cryptogams in the High Arctic: no universal pattern observed in relation to primary
productivity. Pp 11­21 in Jònsdòttir, I.S. (ed), Exploring plant­ecological patterns at
different spatial scales on Svalbard. Unis Publication Series 2006:1
22

Michalet, R., Gandoy, C., Joud, D., Pages, J.P. & Choler, P. 2002: Plant community
composition and biomass on calcareous and siliceous substrates in the northern
French Alps: Comparative effects of soil chemistry and water status. Arctic,
Antarctic and Alpine Research 34: 102­113
Muc, M., Freedman, B. & Svoboda, J. 1989: Vascular plant communities of a polar
oasis at Alexandra Fiord (79 N), Ellesmere Island, Canada. Canadian Journal of
Botany 67: 1126­1136
Murray, D.F. 1997: Regional and local vascular plant diversity in the Arctic. Opera
Botanica 132: 9­18
Palmer, M.W. 1994: Variation in species richness: towards a unification of
hypotheses. Folia Geobotanica & Phytotaxonomica 29: 511–530
Roem, W.J., Berendse, F. 2000: Soil acidity and nutrient supply ratio as possible
factors determining changes in plant species diversity in grassland and heathland
communities. Biological Conservation 92: 151­161
Rohlf, F. 1999: NTSYS­pc. Numerical taxonomy and multivariate analysis system.
Exeter Software, New York
Safronova, I.N. 1980: On the flora of Koteln'y Island (Novosibirsk Islands).
Botanicheskiy Zhurnal 65: 544­551. [In Russian]
Safronova, I.N. 1983: Materials for the flora of Mebel and Hooker Islands (Franz
Joseph Land). Botanicheskiy Zhurnal 68: 513­518. [In Russian]
Safronova, I.N. 1993: On the flora of Bolshevik Island (Severnaya Zemlya
Archipelago). Botanicheskiy Zhurnal 78: 79­84. [In Russian]
Safronova, I.N. & Khodachek, E.A. 1989: On the flora and vegetation of Andrey,
Uyedineniya and Vize islands (northern Arctic Ocean). Botanicheskiy Zhurnal 74:
1003­1011. [In Russian]
Savile, D.B.O. 1959: The botany of Somerset Island, District of Franklin. Canadian
Journal of Botany 37: 959­1002
Savile, D.B.O. 1961: The botany of the northwestern Queen Elizabeth Islands.
Canadian Journal of Botany 39: 909­942
Schofield, W.B. & Cody, W.J. 1955: Botanical investigations on coastal sourthern
Cornwallis Island, Franklin District, NTW. Canadian Field­Naturalist 69: 116­128
Soper, J.H. & Powell, J.M. 1985: Botanical studies in the Lake Hazon region, northern
Ellesmere Island, Northwest Territories, Canada. Canada National Museum of
Natural Sciences Publications in Natural Sciences 5: 1­67
Taylor, D.R., Aarssen, L.W. & Loehle, C. 1990: On the relationship between r/K
selection and environmental carry capacity: a new habitat templet for plant life
history strategies. Oikos 58: 239–250
Theodose , T.A. & Bowman, W.­D. 1997: Nutrient availability, plant abundance, and
species diversity in two alpine tundra communities. Ecology 78: 1861­1872
23

Tilman, D. 1982: Resource Competition and Community Structure. Princeton
University Press, Princeton
Truong, C. 2003: Ecology and genetics of the mountain birch Betula pubescens ssp. tortuosa at the limit
of its geographical range in Northern Sweden: implication for treeline rising in relation to climate
change. Diploma Thesis, University of Neuchâtel, 56 pp
Walker, M.D. 1995: Patterns and causes of arctic plant community diversity. Pp. 1­20
in Chapin, F.S. III & Körner, C. (Eds). Arctic and Alpine Biodiversity. Springer
Verlag, New York
Walker, M. D., Wahren, H. C., Hollister R. D., Henry, G. H. R., Ahlquist, L. E.,
Alatalo, J. M., Bret­Harte, S. M., Calef, M. P., Callaghan, T. V., Carroll, A. B.,
Epstein, H. E., Jónsdóttir, I. S., Klein, J. A., Magnússon, B., Molau, U., Oberbauer,
S. F., Rewa S. P., Robinson, C. H., Shaver, G. R., Suding, K. N., Thompson, , C.
C., Tolvanen, A., Totland, Ø., Turner, P. L., Tweedie, C. E., Webber, P. J. &
Wookey, P. A. 2006: Plant community responses to experimental warming across
the tundra biome. Proceedings of the National Academy of Sciences of the Unites
States of America 103, 1342­1346
Wenche, E., Klanderud, K & Tommelstad, R. 2002: Plant community diversity at
different scales in six localities on Svalbard. Pp 22­40 in Jònsdòttir, I.S. (ed),
Biodiversity in arctic plant communities. Unis Publication Series 2002:3
van der Welle, M.E.W., Vermeulen, P.J., Shaver, G.R. & Berendse, F. 2003: Factors
determining plant species richness in Alaskan arctic tundra. Journal of Vegetation
Science 14: 711­720
Young, S.B. 1971: The vascular flora of St. Lawrence Island with special reference to
floristic zonation in the arctic regions. Contributions from the Gray Herbarium 201:
11­115
24

Appendix 1. Complete species lists from all sites. # means presens in the frames
analysed at the site, * means point frame hit.
1607/1/1; Kapp Thordsen; mesic heath; 8709800, 33X 0511516
Alopecusus borealis Trin.
Arenaria pseudofrigida (Ostenf. & Dahl) Juz.
*Bistorta vivipara (L.) S.F.Gray
Carex fuliginosa Schkuhr subsp. misandra (R.Br.) Nyman
*Carex rupestris All.
Cerastium arcticum Lange coll.
Cochlearia groenlandica L. coll.
#Draba fladnizensis Wulf.
Draba lactea Adams
Draba oxycarpa Sommerf.
Draba subcapitata Simm.
Dryas octopetala L.
#Equisetum arvense L. subsp. alpestre (Wahlenb.) Schönswetter & Elven
*Equisetum variegatum Schleich. ex Weber & Mohr subsp. variegatum
Juncus biglumis L.
Luzula confusa Lindeb.
#Luzula nivalis (Laest.) Spreng.
Minuartia rubella (Wahlenb.) Hiern
Oxyria digyna (L.) Hill
Papaver dahlianum Nordh. subsp. polare (Tolm.) Elven & Ö.Nilsson
Pedicularis dasyantha (Trautv.) Hadac
*Pedicularis hirsuta L.
Poa arctica R.Br.
Potentilla pulchella R.Br. subsp. pulchella
Puccinellia vahliana (Liebm.) Scribn. & Merr.
Ranunculus sulphureus Sol.
Sagina nivalis (Lindbl.) Fr.
*Salix polaris Wahlenb.
Salix reticulata L.
Saxifraga cespitosa L. subsp. cespitosa
#Saxifraga hirculus L. subsp. compacta Hedberg
*Saxifraga oppositifolia L. subsp. oppositifolia
*Silene acaulis (L.) Jacq.
Silene uralensis (Rupr.) Bocquet subsp. arctica (Th.Fr.) Bocquet
*Stellaria longipes Goldie s.l.
1607/1/2; Kapp Thordsen; calcarious fen, 8709858, 33X 0511838
Alopecusus borealis Trin.
*Bistorta vivipara (L.) S.F.Gray
#Carex rupestris All.
*Cerastium arcticum Lange coll.
Cochlearia groenlandica L. coll.
Coptidium lapponicum (L.) Tzvelev
Coptidium x spitsbergense (Hadac) Elven
*Equisetum arvense L. subsp. alpestre (Wahlenb.) Schönswetter & Elven
Equisetum variegatum Schleich. ex Weber & Mohr subsp. variegatum
25

Juncus biglumis L.
Luzula confusa Lindeb.
Luzula nivalis (Laest.) Spreng.
Oxyria digyna (L.) Hill
*Pedicularis dasyantha (Trautv.) Hadac
#Poa alpina (L.) var. vivipara L.
*Poa arctica R.Br.
*Poa pratensis L. alpigena (Fr.) Hiit.
Ranunculus sulphureus Sol.
Ranunculus wilanderi (Nath.) Á.Löve & D.Löve
*Salix polaris Wahlenb.
Salix reticulata L.
Saxifraga cespitosa L. subsp. cespitosa
Saxifraga hieracifolia Waldst. & Kit. ex Willd. subsp. hieracifolia
*Saxifraga hirculus L. subsp. compacta Hedberg
*Saxifraga oppositifolia L. subsp. oppositifolia
Saxifraga svalbardensis Øvstedal
*Stellaria longipes Goldie s.l.
1707/1/2­3; Mayerbreen; old moraine area; 8800608, 33X 0440870
*Cerastium arcticum Lange coll.
Cochlearia groenlandica L. coll.
#Draba alpina L.
*Draba corymbosa R.Br. ex DC.
*Draba lactea Adams
Draba oxycarpa Sommerf.
*Luzula confusa Lindeb.
#Oxyria digyna (L.) Hill
Papaver dahlianum Nordh. subsp. polare (Tolm.) Elven & Ö.Nilsson
*Poa alpina (L.) var. vivipara L.
*Poa arctica R.Br.
*Poa pratensis L. alpigena (Fr.) Hiit.
*Saxifraga cernua L.
*Saxifraga cespitosa L. subsp. cespitosa
*Saxifraga hyperborea R.Br.
*Saxifraga nivalis L.
*Stellaria longipes Goldie s.l.
1707/2/1; Fjortende julibreen; birdcliff; 8786090, 33X 0439992
Bistorta vivipara (L.) S.F.Gray
*Cerastium arcticum Lange coll.
#Cerastium regelii Ostenf.
Chrysosplenium tetrandrum (N. Lund) Th.Fr.
#Cochlearia groenlandica L. coll.
Draba corymbosa R.Br. ex DC.
Draba nivalis Liljebl.
*Draba norvegica Gunn.
Erigeron humilis R.C.Graham
Festuca rubra L. subsp. richardsonii (Hook.) Hultén
Oxyria digyna (L.) Hill
26

*Poa arctica R.Br.
Potentilla hyparctica Malte subsp. hyparctica
Ranunculus pygmaeus Wahlenb.
*Salix polaris Wahlenb.
Saxifraga cernua L.
*Saxifraga cespitosa L. subsp. cespitosa
*Saxifraga hieracifolia Waldst. & Kit. ex Willd. subsp. hieracifolia
Saxifraga nivalis L.
*Saxifraga oppositifolia L. subsp. oppositifolia
*Silene acaulis (L.) Jacq.
*Stellaria longipes Goldie s.l.
Taraxacum brachycera
s Dahlst.
Trisetum spicatum (L.) K.Richt. subsp. spicatum
1707/2/2; Fjortende julibreen; dry heath; 8786090, 33X 0433912
*Cerastium arcticum Lange coll.
Cerastium regelii Ostenf.
Chrysosplenium tetrandrum (N.Lund) Th.Fr.
Cochlearia groenlandica L. coll.
Draba L. sp.
Draba alpina L.
*Draba norvegica Gunn.
Draba oxycarpa Sommerf.
Luzula confusa Lindeb.
*Oxyria digyna (L.) Hill
#Poa alpina L. var. vivipara L.
*Poa arctica R.Br.
*Poa pratensis L. subsp. alpigena (Fr.) Hiit.
Salix polaris Wahlenb.
#Saxifraga cespitosa L. subsp. cespitosa
Saxifraga hieracifolia Waldst. & Kit. ex Willd. subsp. hieracifolia
Saxifraga oppositifolia L. subsp. oppositifolia
Silene acaulis (L.) Jacq.
Stellaria longipes Goldie s.l.
#Taraxacum brachyceras Dahlst.
1807/1/1­2; Biscayarhuken; dry heath; 8864496, 33X 0448709
#Cardamine bellidifolia L. subsp. bellidifolia
*Cerastium arcticum Lange coll.
*Cochlearia groenlandica L. coll.
Draba corymbosa R.Br. ex DC.
#Draba micropetala Hook.
Draba subcapitata Simm.
Luzula confusa Lindeb.
*Luzula nivalis (Laest.) Spreng.
Oxyria digyna (L.) Hill
Papaver dahlianum Nordh. subsp. polare (Tolm.) Elven & Ö.Nilsson
*Poa arctica R.Br.
Potentilla hyparctica Malte subsp. hyparctica
27

Puccinellia phryganodes (Trin.) Scribn. & Merr. subsp. neoarctica
Ranunculus pygmaeus Wahlenb.
Ranunculus sulphureus Sol.
*Salix polaris Wahlenb.
#Saxifraga cespitosa L. subsp. cespitosa
Saxifraga nivalis L.
*Saxifraga rivularis L. subsp. rivularis
*Silene acaulis (L.) Jacq.
#Stellaria longipes Goldie s.l.
Taraxacum arcticum (Trautv.) Dahlst.
Trisetum spicatum (L.) K.Richt. subsp. spicatum
1807/2/1­2; Reinsdyrflya; mesic heath; 8863020, 33X 0478260
#Cardamine bellidifolia L. subsp. bellidifolia
*Cerastium arcticum Lange coll.
*Cerastium regelii Ostenf.
#Cochlearia groenlandica L. coll.
*Draba corymbosa R.Br. ex DC.
Draba fladnizensis Wulf.
Draba oxycarpa Sommerf.
*Draba subcapitata Simm.
Equisetum variegatum Schleich. ex Weber & Mohr subsp. variegatum
#Juncus biglumis L.
Luzula confusa Lindeb.
*Luzula nivalis (Laest.) Spreng.
Minuartia biflora (L.) Schintz. & Thell.
*Oxyria digyna (L.) Hill
Poa alpina L. var. vivipara L.
*Ranunculus pygmaeus Wahlenb.
#Ranunculus sulphureus Sol.
*Salix polaris Wahlenb.
#Saxifraga cernua L.
#Saxifraga cespitosa L. subsp. cespitosa
*Saxifraga oppositifolia L. subsp. oppositifolia
Saxifraga platysepala (Trautv.) Tolm.
#Saxifraga tenuis (Wahlenb.) H.Sm.
*Stellaria longipes Goldie s.l.
1907/1/1; Florabukta, birdcliff; 8888054, 33X 0571330
*Bistorta vivipara (L.) S.F.Gray
Cardamine bellidifolia L. subsp. bellidifolia
*Carex maritima Gunn. coll.
*Cerastium arcticum Lange coll.
Cerastium regelii Ostenf.
*Cochlearia groenlandica L. coll.
*Festuca rubra L. subsp. richardsonii (Hook.) Hultén
Juncus biglumis L.
#Luzula confusa Lindeb.
Luzula nivalis (Laest.) Spreng.
Minuartia biflora (L.) Schintz. & Thell.
28

*Poa arctica R.Br.
Poa pratensis L. alpigena (Fr.) Hiit.
Potentilla hyparctica Malte subsp. hyparctica
*Ranunculus sulphureus Sol.
Salix polaris Wahlenb.
Saxifraga cespitosa L. subsp. cespitosa
Saxifraga oppositifolia L. subsp. oppositifolia
Saxifraga rivularis L. subsp. rivularis
Stellaria humifusa Rottb.
Stellaria longipes Goldie s.l.
Taraxacum arcticum (Trautv.) Dahlst.
1907/1/2; Florabukta; mesic heath; 8887920, 33X 0571369
*Bistorta vivipara (L.) S.F.Gray
Cardamine bellidifolia L. subsp. bellidifolia
Carex maritima Gunn. coll.
*Cerastium arcticum Lange coll.
Cerastium regelii Ostenf.
#Draba lactea Adams
*Festuca rubra L. subsp. richardsonii (Hook.) Hultén
*Luzula confusa Lindeb.
*Luzula nivalis (Laest.) Spreng.
#Minuartia biflora (L.) Schintz. & Thell.
#Minuartia rubella (Wahlenb.) Hiern
Papaver dahlianum Nordh. subsp. polare (Tolm.) Elven & Ö.Nilsson
*Poa arctica R.Br.
Potentilla hyparctica Malte subsp. hyparctica
Ranunculus pygmaeus Wahlenb.
#Ranunculus sulphureus Sol.
*Salix polaris Wahlenb.
*Saxifraga cernua L.
*Saxifraga cespitosa L. subsp. cespitosa
Saxifraga nivalis L.
Saxifraga oppositifolia L. subsp. oppositifolia
1907/1/3; Florabukta; polar desert; 8887344, 33X 0571431
*Cerastium arcticum Lange coll.
#Cochlearia groenlandica L. coll.
#Draba subcapitata Simm.
Juncus biglumis L.
*Luzula nivalis (Laest.) Spreng.
*Minuartia biflora (L.) Schintz. & Thell.
#Minuartia rubella (Wahlenb.) Hiern
#Papaver dahlianum Nordh. subsp. polare (Tolm.) Elven & Ö.Nilsson
#Poa hartzii Gand.
#Potentilla hyparctica Malte subsp. hyparctica
Puccinellia angustata (R.Br.) E.L.Rand & Redfield
*Salix polaris Wahlenb.
#Saxifraga cernua L.
#Saxifraga cespitosa L. subsp. cespitosa
29

#Saxifraga nivalis L.
*Saxifraga oppositifolia L. subsp. oppositifolia
#Saxifraga platysepala (Trautv.) Tolm.
Saxifraga tenuis (Wahlenb.) H.Sm.
2007/1/3; Lady Franklinfjorden; mesic heath; 8901644, 33X 0587985
#Cardamine bellidifolia L. subsp. bellidifolia
#Draba micropetala Hook.
*Luzula confusa Lindeb.
#Papaver dahlianum Nordh. subsp. polare (Tolm.) Elven & Ö.Nilsson
#Saxifraga cernua L.
Saxifraga foliolosa R.Br.
Saxifraga rivularis L. subsp. rivularis
*Stellaria longipes Goldie s.l.
2107/1/3; Ossian Sarsfjellet; mesic heath; 8762902, 33X 0445660
Arenaria pseudofrigida (Ostenf. & Dahl) Juz.
#Bistorta vivipara (L.) S.F.Gray
Carex fuliginosa Schkuhr subsp. misandra (R.Br.) Nyman
Carex nardina Fr. subsp. hepburnii (Boott) Á.Löve, D.Löve & B.M.Kapoor
*Carex rupestris All.
*Cassiope tetragona (L.) D.Don subsp. tetragona
*Dryas octopetala L.
#Luzula nivalis (Laest.) Spreng.
*Salix polaris Wahlenb.
Saxifraga cernua L.
*Saxifraga oppositifolia L. subsp. oppositifolia
*Silene acaulis (L.) Jacq.
Trisetum spicatum (L.) K.Richt. subsp. spicatum
2107/2/1­2; Ossian Sarsfjellet; birdcliff; 8763388, 33X 0445378
Arenaria pseudofrigida (Ostenf. & Dahl) Juz.
*Bistorta vivipara (L.) S.F.Gray
Carex maritima Gunn. coll.
*Cerastium arcticum Lange coll.
*Cerastium regelii Ostenf.
Chrysosplenium tetrandrum (N.Lund) Th.Fr.
Cochlearia groenlandica L. coll.
*Comastoma tenellum (Rottb.) Toyok.
Deschampsia alpina (L.) Roem & Schultes
Draba arctica J.Vahl subsp. arctica
#Draba glabella Pursh
Draba norvegica Gunn.
Dryas octopetala L.
*Erigeron humilis R.C.Graham
*Euphrasia wettsteinii G.Gussarova
*Festuca rubra L. subsp. richardsonii (Hook.) Hultén
Oxyria digyna (L.) Hill
*Poa alpina L. var. vivipara L.
*Poa arctica R.Br.
30

*Poa glauca J.Vahl.
*Poa pratensis L. alpigena (Fr.) Hiit.
*Potentilla hyparctica Malte subsp. hyparctica
Potentilla pulchella R.Br. subsp. pulchella
*Ranunculus affinis R.Br.
*Salix polaris Wahlenb.
*Saxifraga cernua L.
#Saxifraga cespitosa L. subsp. cespitosa
Saxifraga hieracifolia Waldst. & Kit. ex Willd. subsp. hieracifolia
Saxifraga oppositifolia L. subsp. oppositifolia
#Silene acaulis (L.) Jacq.
Silene involucrata (Cham. & Schltdl.) Bocquet subsp. furcata (Raf.) V.V.Petrovsky &
Elven
Silene uralensis (Rupr.) Bocquet subsp. arctica (Th.Fr.) Bocquet
*Stellaria longipes Goldie s.l.
*Taraxacum brachyceras Dahlst.
#Trisetum spicatum (L.) K.Richt. subsp. spicatum
2207/1/1­5; Colesdalen; mesic heath; 8670162, 33X 0502830
*Alopecusus borealis Trin.
*Betula nana L. subsp. tundrarum (Perfil.) Á.Löve & D.Löve
#Bistorta vivipara (L.) S.F.Gray
Campanula rotundifolia L. subsp. gieseckiana (Vest) Mela & Cajander
#Cardamine bellidifolia L. subsp. bellidifolia
Cassiope tetragona (L.) D.Don subsp. tetragona
*Cerastium arcticum Lange coll.
*Cerastium arcticum x regelii
*Dryas octopetala L.
*Equisetum arvense L. subsp. alpestre (Wahlenb.) Schönswetter & Elven
*Euphrasia wettsteinii G.Gussarova
*Festuca rubra L. subsp. richardsonii (Hook.) Hultén
*Hierochloe alpina (Sw.) Roem. & Schult. subsp. alpina
Koeniga islandica L.
*Luzula confusa Lindeb.
Luzula nivalis (Laest.) Spreng.
#Minuartia rubella (Wahlenb.) Hiern
*Pedicularis hirsuta L.
#Poa alpina L. var. vivipara L.
*Poa arctica R.Br.
*Salix polaris Wahlenb.
#Saxifraga hieracifolia Waldst. & Kit. ex Willd. subsp. hieracifolia
Saxifraga svalbardensis Øvstedal
*Stellaria longipes Goldie s.l.
Trisetum spicatum (L.) K.Richt. subsp. spicatum
*Vaccinium uliginosum L. subsp. microphyllum Lange
2207/1/6; Colesdalen; mire; 8670038, 33X 0502916
*Alopecusus borealis Trin.
*Bistorta vivipara (L.) S.F.Gray
*Cerastium arcticum x regelii
31

*Coptidium lapponicum (L.) Tzvelev
Coptidium x spitsbergense (Hadac) Elven
#Dupotia fisheri R.Br.
*Equisetum arvense L. subsp. alpestre (Wahlenb.) Schönswetter & Elven
#Eriophorum scheuchzeri Hoppe subsp. arcticum Novoselova
*Euphrasia wettsteinii G.Gusarova
*Luzula confusa Lindeb.
#Luzula nivalis (Laest.) Spreng.
#Pedicularis hirsuta L.
*Poa arctica R.Br.
*Ranunculus hyperboreus Rottb. subsp. arnellii Scheutz
*Salix polaris Wahlenb.
#Saxifraga foliolosa R.Br.
#Saxifraga hieracifolia Waldst. & Kit. ex Willd subsp. hieracifolia
*Saxifraga svalbardensis Øvstedal
32

Life in the Arctic – a struggle for survival?
Simone Lang 1 , Merete Wiken Dees 2 and Kathrin Bockmühl 3
1 Systems Ecology, Faculteit der Aard­ en Levenswetensschappen, Vrije Universiteit Amsterdam, De
Boelelaan 1085, NL­1081 HV Amsterdam, The Netherlands, simone.lang@ecology.falw.vu.nl,
2 Institutt for plante og miljøvitenskap, Universitetet for miljø og biovitenskap, NO­1432 Ås, Norway,
mwdees@student.umb.no, 3 Department of biology, University of Bergen, Allégaten 41, NO­5001
Bergen, Norway, kathrin.bockmuhl@student.uib.no
Abstract
Are widespread plants in the Arctic widespread due to their wide ecological amplitude
or due to a widespread habitat? To address this question, ecological amplitudes of
plant species across the Arctic were investigated. Vascular plant species composition
and abiotic factors were recorded in 88 plots at 29 sites in 10 different regions on
Svalbard. For each species the ecological amplitude was recorded according to Raups
study carried out at Mesters Vig on North East Greenland in 1969. The ecological
amplitudes of plant species occurring on both Svalbard and Mesters Vig district were
compared. Some differences were found, but this was not the case when Svalbard was
compared to the entire Greenland. pH and bioclimatic zones proved to be the driving
factors for plant distribution on Svalbard. The thermophilic annual Euphrasia
wettsteinii seems to be a sensitive indicator for current climate change scenarios. Our
study showed that widespread taxa do not automatically possess wide amplitudes on
Svalbard and vice versa.
Introduction
Climate change at high latitudes is predicted to be greater and more rapid than in any
other region on Earth. According to the (ACIA 2006) not only will temperatures
change (Przybylak 2007) but also precipitation is expected to increase in wide regions
in the Arctic. Recent observations and experimental studies have supplied evidence
that arctic plants and ecosystems react substantially to rising temperatures (Walker et
al. 2006). Increasing temperatures and rainfall might induce permafrost melting,
leading to a changed hydrology and unstable substrates.
Arctic plants are exposed to extreme amplitudes of environmental factors: early snow
and frost, drought, wind abrasion and extreme temperatures on open patches (Billings
& Mooney 1968). Slope instability can occur as stochastic mass flows as well as slow
continuous process, depending upon the geomorphic circumstances. Low temperature
and short growing seasons result in vegetation dominated by long­lived perennials
that reproduce through vegetative growth, while annuals are rare (Billings & Mooney
1968, Savile 1972). The flora of Svalbard consists of about 170 flowering plant
species and only two of those are annual species: the cold tolerant Koenigia islandica
and the rare, thermophilic hemiparasitic species Euphrasia wettsteinii (Brochmann &
Steen 1999). Euphrasia wettsteinii occurs at its northern limit on Svalbard.
Thermophilic annuals may be good indicators for climate change due to their short
33

esponse curve. With rising temperatures an increase in population size can be
expected.
What are the limits for plant distribution in the Arctic? Various biotic and abiotic
factors such as temperature and moisture regimes, nutrient supply, competition,
dispersal limitation, seedling survival, to name only a few, restrict plant life in the
Arctic. Yet can we conclude a ‘struggle for survival’ or are we observing rather well­
adapted plant species being able to withstand extremely harsh conditions? Alsos et al.
(2007) showed that dispersal of plants is not a limiting factor in the Arctic. A variety
of studies on ecological amplitudes of plant species has been conducted in various
parts of the Arctic and elsewhere (Coudun & Gegout 2005, Saetersdal & Birks 1997).
Karlsen & Elvebakk (2003) grouped species in Greenland according to their minimum
temperature and soil pH preferences. Are ecological amplitudes derived from such
studies applicable across the Arctic? Ellenbergs (Ellenberg et al. 1991) indicator
values, being developed for middle Europe including the Alps, needed to be modified
when applied to other regions such as England (Hill et al. 2000). The same situation
may arise when comparing plant taxa occurring at different locations in the Arctic.
Raup (1969) conducted a study of plant behaviour at Mester Vig, North East
Greenland where he investigated the effects of disturbance, moisture and total plant
cover on plant distribution. No attempt has been made to replicate his study in other
parts of the Arctic comparing the ecological amplitudes of the same plant taxa. If we
are to understand, generalise and model plant responses to climate change we will
need to identify their ecological amplitudes in response to numerous biotic and abiotic
factors at a multitude of locations in the Arctic.
In this study we attempt to assess the ecological amplitudes of arctic vascular plant
species on Svalbard in terms of their site factors such as substrate, moisture,
disturbance and temperature range. Svalbard is especially suitable for this study since
it covers three Arctic bioclimatic zones from polar deserts in zone A with extremely
low productivity to relatively productive tundra in zone C. In addition, a great variety
of substrates and geomorphological processes facilitates a recording of species at their
ecological optimum and also at their extremes (Jónsdóttir 2005). We furthermore
compare the amplitudes of the species found both on Svalbard and the Mesters Vig
District in Greenland according to Raup’s recording method.
In our study we address the following questions:
1. Are widespread plants in the Arctic widespread due to their wide ecological
amplitude or due to a widespread habitat?
1. Do the same species from Svalbard and Greenland on similar landscapes differ
in their ecological amplitude?
2. Will the thermophilic species Euphrasia wettsteinii expand its range if the
mean July isotherme increases by 1 degree Kelvin?
Methods
Study area
Svalbard is an archipelago situated in the High Arctic at 74­/81°N and 10­/30°E. Ten
localities along the north­west coast of Spitsbergen and the west­coast of
34

Nordaustlandet (78­80°N) were studied during the period of 16.­22. July 2007 (Figure
1). The localities covered the three bioclimatic zones of Svalbard according to
Elvebakk (Elvebakk 1999, Elvebakk 2005) including a wide variety of habitats. Two
of the sites were located in zone A, four were located in zone B and four in zone C
(Table 1). Zone A, B and C represent the arctic polar desert zone, the northern arctic
tundra zone and the middle arctic tundra zone, respectively.
Table 1. Bioclimatic zones, general vegetation and bedrock type for the regions in Svalbard
assessed in the study.
Region Bioclimatic
zone
Site
(# of
plots)
General
vegetation
35
Site
description
Bedrock pH1
Lady
1 (3) 5,25
Franklinfjorden A 2 (3) Luzula confusa Polar desert Acidic 5,27
3 (3)
5,47
1 (3)
5,45
Florabukta B (A)2
Mesic Luzula Underneath Alkaline
nivalis Bird cliff circumneutral
2 (3) Next to
Bird cliff
6,21
3 (3)
Polar desert
7,90
1 (5) Young
6,53
Mayerbreen B
Mesic Luzula moraine Acidic
2 (3) confusa Old moraine
6,73
3 (4)
6,93
Fjortende
1 (3) Bird cliff Acidic 7,16
Julibreen B 2 (3)
7,06
1 (3) Hill, north
5,46
Biscayarhuken B
Mesic Luzula facing slope Acidic
2 (2) confusa Hill, south
facing slope
5,46
1 (3) Poa alpina Snow bed Sligthly acidic 6,10
Reinsdyrsflya B 2 (3)
6,51
1 (3) Mesic Dryas Heath Weakly acidic 7,33
Kapp Thordsen C 2 (3) Calcarious fens 6,48
1 (3)
8,31
Kongsbreen C
Cassiope Very young Acidic to
tetragona moraine alkaline
2 (3) Intermediate
moraine
cirucumneutral 8,41
3 (3)
Established
vegetation
7,92
1 (2) Luzula confusa Underneath Acidic to 7,78
Ossian Sars C 2 (2)
bird cliff alkaline
cirucumneutral
7,99
1 (4) 5,84
C 2 (4) Dryas­ Heath
5,75
3 (4) Cassiope
5,72
Colesdalen
4 (4) tetragona
Weakly acidic 5,57
5 (2)
5,41
6 (4)
Fen
5,71
1
pH was measured in this study
2
At the border between A and B

N
Biscayarhuken
Florabukta
Reinsdyrsflya
Mayerbreen
Fjortende Julibreen
Ossian Sars
Kongsbreen
Kapp Thordsen
Colesdalen
Figure 1. Map of Svalbard showing the sampling sites.
36
Lady Franklinfjorden

The percentage area of different substrate types of Svalbard was estimated by means
of the preliminary substrate map of Svalbard (Arnesen, personal communication)
using GIPM Version 2.2.1.1. The pixels of each substrate category were counted and
expressed as percentage area in relation to the pixel sum of entire Svalbard. The
dominant substrate type is weakly acidic and covers 14.1% of entire Svalbard. Acidic
substrates cover 6.1%, followed by the alkaline circumneutral type with 5.4%. The
remaining substrate consists of sediments (3.6%), saline steppe substrate (0.1%) and
undefined substrate (0.2%). Svalbard is dominated by glaciers occupying 67.5% of
the surface area.
Meteorological data
Data from the meteorological station at Svalbard Airport (MET 2007) were used to
calculate the increase in mean summer temperatures for the last ten years (1997­
2007). Figure 2 shows a simplified model of the temperature data using a linear
regression. In the period 1997­2007 mean summer temperatures increased with a rate
of 0.16 K per year. As a comparison the mean summer temperature of the last normal
period (1961­1990) is included in the diagram.
Degrees Celsius
7
6
5
4
3
2
1
0
1997
1998
Mean summer temperature at Svalbard Airport
1999
2000
2001
2002
2003
2004
37
2005
2006
2007
Mean temperatures
Regression model
Mean summer temperature
1961­1990
Figure 2. Mean summer (June, July and august) temperature at Svalbard airport (R 2 = 0.52).
Study species
The annual thermophilic Euphrasia wettsteinii G.Gusarova belongs to the family
Scrophulariaceae. The species was earlier treated as E. frigida Pugsley (PAF 2007).
Euphrasia wettsteinii is a root hemiparasite, but it is still unclear which hosts the
Svalbard­populations are parasitising (Alsos et al. 2004). Some hemiparasites are
reported with only one host. However, E. wettsteinii is found physically connected to
a variety of hosts or it may grow unconnected to a host (Seel & Press 1993).
Euphrasia wettsteinii is mainly self­pollinating. The plants are found in closed plant
communites, therefore openings in the cover are needed to expose the soil and provide
a seed bed. Some disturbance which causes small open patches in the growth sites is
needed for establishment of the seeds (Molau 1993, Seel & Press 1993). Reproducing
by seeds may be of advantage to survive unfavourable winter times, but a long series

of cold summers can be terminal for a population of Euphrasia depending on their
seed longevity which may only last for a few years.
Euphrasia wettsteinii was first recorded near the hot spring in Bockfjord (Rønning
1961). In 1998 one patch of E. wettsteinii was discovered in Colesdalen (Alsos et al.
2004) and one at Ossian Sars five years later (Alsos). In 2002 a detailed investigation
of the flora in Colesdalen was carried out and nine patches of E. wettsteinii were
discovered. All the patches except two were less than 2 m 2 . E. wettsteinii is assigned
to the Red List category endangered (Bakken et al. 2006).
Vegetation data
Two to six sites were selected within each of the ten localities and a total of 88 plots
were chosen among sites. The plots were randomly chosen for unbiased application of
plot sampling. Exceptions from the random selection of plots were the Euphrasia­
sites on Ossian Sars and Colesdalen. The plots in these two regions were subjectively
selected to include patches of E. wettsteinii. The size of the entire patch of E.
wettsteinii at Ossian Sars was 1m x 2m. Thus only two plots with and two without E.
wettsteinii were recorded. Ten plots with E. wettsteinii and ten excluding the species
were assessed in Colesdalen. The point intercept method was used to record the
vascular species within each plot (Bråthen & Hagberg 2004). At each site, 50 cm x 50
cm plots with 25 points of equal intercepts of 10 cm were used. All vascular plants
touched by the needle at each point were recorded. Each taxon was counted only once
at any one point. Two to four replicate plots per site were recorded to obtain an
adequate representation of the community. Additional species in the plot were
recorded and counted as 0.5 hits. The total vegetation cover in the plots was estimated
in percent vascular plant cover, bryophyte cover, lichen cover, bare ground/rocks and
cryptogram crust cover. The black and white cryptogam crust consisted of a mixture
of cyanobacteria, white crustose lichens and a few minute bryophytes. Species names
of the vascular plants followed PAF (2007).
A distribution estimate for each species was calculated to test whether species are
truly widespread on Svalbard. Svalbard is divided in 26 districts. For the species
encountered in the field a count per district was carried out using the Svalbard
database (Alsos, personal communication). The Svalbard database includes all
specimens recorded in Norwegian herbaria as well as a list of species for particular
sites. The count per district was used as a distribution estimate of each species. Poa
alpina var. vivipara and P. alpina var. alpina were combined to P. alpina. The
northernmost distribution limit was based on P. alpina var. vivipara which extends
further north. In the district count of Poa arctica both P. arctica ssp. caespitans and
P. arctica var. vivipara were included.
Abiotic data
Abiotic data were assessed per plot. Organic soil depth was measured and divided into
the three categories of low (0­2 cm), medium (2­5 cm) and high (>5 cm) organic soil
depth. Permafrost measurements could not be consistently taken at all sites and were
therefore not included in the analysis. Soil temperature was measured at 10 cm soil
depth using a handheld digital thermometer. Slope and exposure of the plots were
assessed using a compass. For pH measurements soil samples were taken from the
uppermost 5 cm of the soil regardless of organic or mineral origin since this layer
constitutes the main rooting zone for the plants. The samples were kept cool and
38

processed within 24 hours. Equal amounts in volume of both soil and deionised water
were mixed in a flask. After repeated shaking for 1 hour, pH was measured by using
an electronic pH­meter. The pH values were divided into three categories of low (<
5.2), medium (5.2­7.0) and high (> 7.0) pH. This division corresponds to Karlsen and
Elvebakk (2003). The definition of substrates follows the preliminary substrate map
of Svalbard (Arnesen, personal communication). Raup did not include pH­
measurements in his study at the Mesters Vig district. Data from Karlsen & Elvebakk
(2003) were used for the Svalbard ­ Greenland comparison.
The soil moisture within each plot was assessed according to Raup (1969) by feeling
the soil with the fingers. Following categories were used:
1. low (dry, no feeling of dampness to the fingers)
2. medium (moist, wet the fingers when handled)
3. high (wet, water can be squeezed from the soil)
4. very high (wet, water is freely dripping from soil sample)
The disturbance gradient was separated into two categories established by Raup
(1969). He distinguished between frost disturbance (sorted or non­sorted circles) and
non­frost disturbance (erosion, slope instability). Although being a frost­initiated
process, Raup included solifluction into the non­frost disturbance processes since
plant roots are located in the upper layer of the sliding soil and thus not disturbed by
the slow soil movement. The categories for both frost and non­frost disturbance were
divided into:
1. stable (no visual disturbance, no bare ground)
2. medium (bare ground visible indicating cryoturbation and enabling erosion)
3. unstable (bare ground dominating, active cryoturbation or erosion, no
consolidation of soil recognisable)
Statistical analysis
To find the main abiotic factors determining species distribution, a DCA, CA and
CCA ordination were carried out using Canoco (Canoco for Windows 4.5, 1999). Hits
per species obtained by the point intercept method were used as the species factor.
Environmental factors included pH, soil temperature, organic soil depth, moisture,
frost and non­frost disturbance and total vascular plant cover. For ordination the
values and not the categories for pH and organic soil depth were used. Since soil
temperature was measured only once at each site being strongly dependant on current
weather conditions, bioclimatic zones A to C were used to replace soil temperature in
a second ordination. Bioclimatic zones are defined as subdivisions of the Arctic
reflecting climatic gradients (Elvebakk 1999). They integrate climatic measurements
over a longer time period thus greatly improving the comparison between the different
localities used in this study. The dummy variables for bioclimatic zones were
numbered 1 to 3 for zone A to C according to rising temperatures. Subdivision of
arctic reflecting climatic gradients.
We assumed that the species was temperature dependant so we wanted to test whether
soil temperature could be a factor differing between plots. Differences in soil
temperature for the plots at Colesdalen where E. wettsteinii was growing compared to
the plots excluding the species were tested by carrying out a t­test using S­Plus® 6.2
for Windows (Insightful Corp. Seattle, WA, USA, 2003).
39

Results
Ecological amplitudes of vascular plant species on Svalbard
The DCA ordination revealed that the length of the gradient was larger than four.
Thus a PCA based on species showing a linear response along a gradient could not be
carried out. The Monte Carlo Permutation test obtained from a CCA proved that soil
temperature was not significant (p = 0.078) whereas bioclimatic zones used as a
temperature proxy showed high significance (p = 0.002). The variable bioclimatic
zone was consequently used in the following analysis. The Monte Carlo Permutation
test showed that all environmental variables except frost and non­frost were
significant (Table 2).
Table 2. Significance values of the environmental variables in the Monte Carlo Permutation
test.
p­value F­ratio
pH 0.002 3.93
Bioclimatic zone 0.002 3.89
Vascular plant cover 0.002 2.84
Organic soil depth 0.002 2.51
Moisture 0.024 1.84
Frost disturbance 0.158 1.26
Non­frost disturbance 0.362 1.04
The CA correlation matrix revealed that the bioclimatic zone was weakly correlated
with all other factors except pH (Table 3).
Table 3. CA correlation matrix based on hits per species. Environmental variables used are
bioclimatic zones, organic soil depth, moisture, non­frost disturbance, frost
disturbance, pH and total vascular plant cover.
Bioclimatic Organic Soil Non­frost Frost pH Vascular
zone
soil depth moisture disturbace disturbance
Plant Cover
Bioclimatic zone 1
Organic soil depth 0.2106 1
Soil moisture
Non­frost
0.2990 0.1529 1
disturbance 0.2093 ­0.1290 ­0.0554 1
Frost disturbance 0.1469 0.1441 0.1514 0.0524 1
pH
Vascular plant
0.0965 0.0820 ­0.2484 ­0.1482 ­0.0170 1
cover 0.3060 ­0.0954 0.0203 0.0791 ­0.1125 0.1192 1
Figure 3 illustrates the driving factors of plant distribution on Svalbard. The first and
second axis explained 7.8% and 7% of the variation, respectively.
The two non­correlated factors pH and bioclimatic zone were regarded as the main
factors for plant distribution on Svalbard. Species were sorted according to their pH
ranges and each pH range subsequently divided into four groups following their
northernmost (scattered or frequent) occurrence in the bioclimatic zones A – D (PAF
2007). Thus both the main distribution limit and the temperature limitation were
combined in one table (Table 4). Bioclimatic zone D does not occur on Svalbard. It is
included in the table to be consistent regarding the northernmost main plant
40

distribution. The complete results for the ecological amplitudes of 67 species on
Svalbard are found in Appendix 1.
As can be seen in Table 4, the smallest group consists of plants with the widest
amplitude in pH range, most of them with a northernmost occurrence in bioclimatic
zone A.
Twenty­eight species show a medium range for pH values medium to high whereas
only one plant, Luzula confusa, has a medium range for pH levels from low to
medium. The emphasis of the distribution limit focuses on the northernmost
bioclimatic zone A.
No taxon can be found for the group with narrow range and low pH. Twenty­one
species show a narrow range for medium pH. Thirteen taxa are found in the group
with preferences for high pH only. In all groups a tendency can be seen that the
category including most taxa for a given pH range displays its northernmost main
distribution limit in bioclimatic zone A.
Figure 3. CA ordination of hits per species (67 in total) in 88 plots, located on Svalbard, in
relation to environmental variables. Stars indicate significant variables. Sample and
species labels were omitted in the figure. First and second axis explained 7.8% and
7%, respectively.
41

A potential substrate area could be calculated for each species knowing the species­
specific pH range by summing the percentage area of acidic, weakly acidic and/or
alkaline circumneutral substrate occurring on Svalbard. These were set in relation to
the distribution estimate of each species (Figure 4).
The number of species occurring on all substrates is low and corresponds to the
species with the widest pH range from acidic to alkaline circumneutral substrate
material (Table 4). All of the species in this group show a wide distribution. On all
substrate type combinations more than half of the species of each group are
distributed over ten districts, except the acidic ­ weakly acidic with Luzula confusa as
the only group member.
42

High pH amplitude
Medium pH amplitude
Narrow pH amplitude
Table 4. pH tolerance of 67 plant species on Svalbard sorted by bioclimatic zones.
Low to high pH
zone A zone B zone C zone D
Cerastium arcticum Salix polaris None None
Poa arctica
Saxifraga cespitosa ssp. cespitosa
Low to medium pH
zone A zone B zone C zone D
Luzula confusa None None None
medium to high pH
zone A zone B zone C zone D
Bistorta vivipara Equisetum arvense ssp. alpestre
Equisetum variegatum ssp.
Carex maritima
Cerastium regelii
variegatum Carex rupestris
Cochlearia groenlandica Festuca rubra ssp. richardsonii Draba norvegica
Luzula nivalis Poa alpina Dryas octopetala
Micranthes nivalis Poa pratensis ssp. alpigena Euphrasia wettsteinii
Micranthes hieracifolia ssp.
Minuartia rubella
hieracifolia
Oxyria digyna Minuartia biflora
Papaver dahlianum ssp.polare Pedicularis dasyantha
Saxifraga cernua Taraxacum brachyceras
Saxifraga hirculus ssp. compacta
Saxifraga oppositifolia ssp.
oppositifolia Unknown zone:
Silene acaulis Cerastium regelii x arcticum
Stellaria crassipes
Low pH
None
Medium pH
zone B zone B zone C zone D
Alopecurus borealis Dupontia fisheri Coptidium lapponicum
Vaccinium uliginosum ssp.
Betula nana ssp. tundrarum
Cardamine bellidifolia ssp. bellidifolia Eriophorum scheuchzeri coll. microphyllum
Draba alpina Hierochloe alpina ssp. alpina
Draba corymbosa Pedicularis hirsuta
Ranunculus hyperboreus ssp.
Draba lactea
arnellii
Draba micropetala Saxifraga rivularis ssp. rivularis
Juncus biglumis
Micranthes foliolosa
Micranthes tenuis
Ranunculus pygmaeus
Ranunculus sulphureus
Saxifraga hyperborea
high pH
zone A zone B zone C
Cassiope tetragona ssp.
zone D
Draba subcapitata Draba glabella
tetragona Comastoma tenellum
Phippsia algida Poa glauca Draba fladnizensis
Poa abbreviata ssp. abbreviata Trisetum spicatum ssp. spicatum Erigeron humilis
Potentilla hyparctica ssp. hyparctica Ranunculus arcticus
Saxifraga platysepala
43

no. of districts
30
25
20
15
10
5
0
alkaline
circumneutral (5.4%)
weakly acidic (14.1%)
increasing substrate area defined by species pH range
44
Weakly acidic – alkaline
circumneutral (19.5%)
acidic – weakly
acidic (20.2%)
acidic – alkaline
Circumneutral Circumneutral (25.6%)
Figure 4. Occurrence of species in districts in relation to sum of pH defined substrate area.
Comparison of ecological amplitudes of selected species from Svalbard and
Greenland, with special emphasis on the Mesters Vig district, NE Greenland
During the fieldwork on Svalbard 67 species were observed in total, of which 45 had
previously also been studied in Greenland. The ecological amplitudes of the species
present on both Svalbard and at Mesters Vig were compared. Different ecological
amplitudes were found for ten species (Table 5). For the complete list and ecological
amplitude of plant taxa occurring on both Svalbard and Mesters Vig, NE Greenland,
see Appendix 2.
Betula nana ssp. tundrarum on Svalbard was observed in plots with 80­90 % total
vascular plant cover. At Mesters Vig, the species was found to be growing at all
categories from low to high density of total vascular plant cover. In addition, B. nana
showed a broader amplitude on Svalbard regarding non­frost disturbance than at
Mesters Vig. Cardamine bellidifolia ssp. bellidifolia on Svalbard was found to grow
on stable to medium stable ground with respect to frost disturbance whereas at
Mesters Vig the species only occurred on stable substrate. On Svalbard, Cochlearia
groenlandica displayed a broader amplitude regarding total vascular plant cover than
the species at Mesters Vig. Minuartia rubella and Poa abbreviata ssp. abbreviata
were both growing on medium moist soil on Svalbard, while they occurred on soil
with low moisture content at Mesters Vig. Phippsia algida was found on Svalbard on
stable ground with frost disturbance whereas it occurred on medium disturbed
substrate at Mesters Vig. Poa pratensis ssp. alpigena and Saxifraga cespitosa ssp.
cespitosa possessed on Svalbard a broader amplitude regarding the vascular plant
cover than at Mesters Vig. Micranthes foliolosa ssp. foliolosa was present in this
study at a single site occurring with 50­60 % total vascular plant cover. At Mesters

Vig, the species was recorded in places with 1­20 % vascular plant cover. Micranthes
hieracifolia ssp. hieracifolia on Svalbard showed a lower tolerance to frost
disturbance and a slightly higher tolerance with respect to non­frost­disturbance
compared to the Mesters Vig district.
Table 5. Ecological amplitude of ten selected species occurring on both Svalbard (blue) and
Greenland (red). The thin lines represent not observed but suggested parts of the
respective amplitudes. The discussed differences are marked in light blue.
Species
1­10
11­20
21­30
total plant cover moisture frost dist.
31­40
41­50
51­60
61­70
Regarding the pH, Karlsen & Elvebakk (2003) did not find any species growing
solely on acidic substrates which corresponds to our findings. Table 6 shows a
comparison of the results for species being present in both the Karlsen & Elvebakk
study and ours. The assigned pH range for species on Svalbard is narrower for Betula
nana ssp. tundrarum, Draba alpina, Draba glabella and Eriophorum scheuchzeri
coll. Different pH ranges were found for the species Cassiope tetragona ssp.
45
71­80
81­90
low
medium
high
stable
medium
unstable
stable
non­frost
dist.
Betula nana ssp. tundrarum 2 1
Cardamine bellidifolia ssp.
Bellidifolia 9 6
Cochlearia groenlandica 8 5
Micranthes foliolosa 1 1
Micranthes hieracifolia ssp.
Hieracifolia 5 3
Minuartia rubella 4 4
Phippsia algida 1 1
Poa abbreviata ssp.
Abbreviate 1 1
Poa pratensis ssp. alpigena 6 5
Saxifraga cespitosa ssp.
Cespitosa 24 12
medium
unstable
presence in plots (total: 88)
presence per site (total: 29)

tetragona, Erigeron humilis, Minuartia biflora and Trisetum spicatum ssp. spicatum.
These species occurred at one site with a pH greater than 7.
Table 6. Comparison of pH ranges (C: calcareous (pH >7); N: neutral (pH 5.2­7.0); A: acidic
(pH

Table 7. Presents of possible host species of Euphrasia wettsteinii in the plots at Ossian Sars
and Colesdalen.
Species Plot with
Euphrasia
(n = 2)
47
Ossian Sars Colesdalen
Plot without
Euphrasia
(n = 2)
Plot with
Euphrasia
(n = 10)
Plot without
Euphrasia
(n = 10)
Bistorta vivipara 100% 100% 100% 90%
Festuca rubra ssp. richardsonii 100% 50% 100% 80%
Poa arctica 50% ­ 20% 10%
Salix polaris ­ 50% 90% 70%
Equisetum arvense ssp. alpestre ­ ­ 100% 100%
Stellaria crassipes 50% 50% 90% 80%
Trisetum spicatum ssp. spicatum 100% ­ 30% 30%
Luzula confusa ­ ­ ­
Alopecurus borealis ­ ­ 80% 80%
Mean soil temperature (10 cm depth) 8,5°C 7,8°C 6,1°C 5,0°C
Discussion
Comparison of ecological amplitudes of selected species from Svalbard and
Greenland, with special emphasis on the Mesters Vig district, NE Greenland
The range of this study limited our observations of the habitat types in which species
might be able to grow. Species with narrow ecological amplitudes might show
broader amplitudes than recorded here. To avoid misinterpretation or generalization of
a species’ ecological amplitude we focused in this comparison on species with
broader ecological amplitudes or different non­congruent amplitudes vis­à­vis
Mesters Vig.
Ten of the observed species in Svalbard showed differences in their ecological
amplitudes in comparison to the same species found at Mesters Vig. However,
comparisons with the distribution map of Svalbard (Elvebakk 2005, Hultén & Fries
1986) revealed that all but Poa pratensis ssp. alpigena and Saxifraga cespitosa ssp.
cespitosa, were widespread beyond the areas we visited. Poa pratensis ssp. alpigena
and Saxifraga cespitosa ssp. cespitosa displayed wider ecological amplitudes on
Svalbard than at Mesters Vig. Cardamine bellidifolia ssp. bellidifolia and Micranthes
hieracifolia ssp. hieracifolia showed broader ecological amplitudes on Svalbard than
at Mesters Vig in terms of their tolerance to frost disturbance. Cochlearia
groenlandica and Micranthes foliolosa were associated with a wider range of total
vascular plant cover being wider on Svalbard. However, literature research (Böcher
1963, Gelting 1934, Raup 1965, Raup 1969, Sørensen 1933) provided evidence that
no difference in ecological amplitudes exists for a given taxon present on both
Svalbard and Greenland when comparing all of Greenland (Böcher 1963, Raup 1965).
The main difference found was the plants tolerance to frost disturbance. The species
at Mesters Vig seem to have wider amplitudes regarding frost disturbance compared

with the species occurring on Svalbard. Although Mesters Vig is located slightly more
south than Svalbard, frost disturbance is the major geomorphologic process
determining plant life there. Frost cracking and heaving are widespread processes
operating in the Mesters Vig (Washburn 1969). Most of the localities we sampled on
Svalbard did not show a high degree of frost disturbance. The process of frost
disturbance depends on topography, rock material, temperature, moisture and snow
cover. Time may be of importance when frost patterns develop very slowly
(Washburn 1973). Old exposed landscape surfaces where frost features are
predominant were not assessed in this study explaining part of the differences in the
results.
The species at Mesters Vig tend to exhibit a wider tolerance on the moisture gradient
compared to Svalbard. This may be due to the limited dataset of this study. Soil
moisture measurements were conducted only once with most measurements showing
medium soil moisture. Minuartia rubella and Poa abbreviata ssp. abbreviata were
growing on medium moist substrate on Svalbard in contrast to the Mesters Vig district
where both species were present at low soil moisture. At Florabukta where the plots
were recorded, precipitation led to overestimation of soil moisture. If low soil
moisture contents is typical at this location, an underestimation of the moisture
amplitude for all species present at this site is likely.
For Betula nana ssp. tundrarum broader amplitudes with respect to total plant cover
were observed on Greenland compared to Svalbard. Since this species is rare on
Svalbard, these differences may be due simply to limited observations.
Phippsia algida was found on Svalbard on stable with respect to frost disturbance in
contrast to the Mesters Vig district where the species was growing on medium
disturbed substrates. This difference can partly be due to limited observations on
Svalbard. In the Mesters Vig district it is possible that stable, non­frost disturbed
ground is rarely present.
Owing to the fact that its taxonomy has been heavily revised during past years,
Cerastium arcticum was not compared since we cannot be certain what we found on
Svalbard is the same taxon as at Mesters Vig. Morphological variation is high within
the complex, and the species are not clearly distinguishable (Hagen 2001). Thus it
remains unclear to which species Raup referred in his study at Mesters Vig.
Different pH ranges for the species Cassiope tetragona ssp. tetragona, Erigeron
humilis, Minuartia biflora and Trisetum spicatum ssp. spicatum were observed at one
site with a pH clearly higher than 7, in contrast to the report by Karlsson & Elvebakk
(2003) of acidic to weakly acidic pH ranges for the same taxa. However, Cassiope
tetragona ssp. tetragona is known to grow at sites with limestone influence in sub­
arctic habitats in northern Sweden (Lid et al. 2005).
The pH range recorded in this study seems to be too narrow for the species Betula
nana ssp. tundrarum, Draba alpina, Draba glabella and Eriophorum scheuchzeri
coll. when comparing to Karlsson & Elvebakk (2003). This may be due to the fact that
these species were recorded at one site only. These differences might disappear when
applying a larger dataset.
48

Taking all of Greenland into account taxa occurring both on Svalbard and Greenland
tend to show the same overall ecological amplitude regardless of low or high arctic
environment. However, further studies are necessary to verify this statement. As yet,
generalisations of ecological amplitudes of plant taxa for the entire Arctic should be
considered with care.
Ecological amplitudes of vascular plant species on Svalbard
The results show that the majority of the species over all pH ranges exhibit their main
distribution limit in bioclimatic zone A. Regarding pH amplitudes very few plants
(five species) seem to be able to grow at low pH (< 5.2) values. Most prefer medium
to high pH substrates (62 species). This reflects the substrate availability on Svalbard
with weakly acidic substrates (pH 5.2­7.0) covering the largest area.
Cerastium arcticum, Poa arctica, Saxifraga cespitosa ssp. cespitosa and Salix polaris
are truly widespread species with a wide ecological amplitude in terms of both pH and
temperature. In addition, they show a wide tolerance concerning total plant cover (10­
90%) and a medium range for disturbance and moisture (s. Appendix 1). Species
which are restricted to bioclimatic zone C or D and exhibit a narrow pH amplitude are
less widespread.
More than half of the species in each substrate group except the acidic and weakly
acidic substrate group are widespread on Svalbard. All three substrate types occur in
each bioclimatic zone on Svalbard. Thus the substrate type does not appear to limit
distribution of species on Svalbard.
Luzula confusa and Poa arctica, although widespread on the species distribution map
of Svalbard, received relatively low values in the district count per species. This may
be due to an underrepresentation of the taxa in the Svalbard database which served as
a basis for the district count.
Ecological amplitudes of a few selected species were compared to the vegetation type
map (Arnesen, personal communication), species distribution map and substrate map
of Svalbard (Arnesen, personal communication) to verify our findings. Our study
shows that Luzula confusa prefers substrates of low to medium pH. The substrate on
Nordaustlandet, where the Luzula confusa vegetation type is dominant, consists
mainly of acidic substrate types reflecting the species preferences. In addition, the
distribution map shows that Luzula confusa doesn’t occur in the area of alkaline
circumneutral substrates but is present on acidic and weakly acidic substrates. The
medium to high pH values which were recorded for Papaver dahlianum ssp. polare,
are congruent with its occurrence in the Papaver polar desert where partly alkaline
and weakly acidic substrates are dominant. The mesic Luzula nivalis vegetation type
of bioclimatic zone B relates to weakly acidic and alkaline substrates and the assigned
pH range of Luzula nivalis. The distribution of the Poa alpina snow beds corresponds
to a substrate higher in pH which is expressed in the species pH tolerance of medium
to high pH values. On the species distribution map, Papaver dahlianum ssp. polare,
Luzula nivalis and Poa alpina are mainly distributed in areas with weakly acidic to
alkaline circumneutral substrates matching their ecological pH amplitudes. The
Cassiope tetragona ssp. tetragona vegetation type, occurring on weakly acidic
substrates, is less well expressed in our data where a narrow pH range of high pH
values was recorded for C. tetragona ssp. tetragona. However, the species distribution
49

and substrate map show a clear occurrence of Cassiope tetragona ssp. tetragona on
alkaline circumneutral substrates. Overall, the data from this study are consistent
despite the relatively small dataset.
We concentrated on the variables pH and bioclimatic zone since zone was weakly
correlated to all other factors except pH. Accordingly, species categories could have
been devised in relation to the variables being significant in the Monte Carlo
Permutation Test (s. Table 2) such as total vascular plant cover, organic soil depth and
moisture. Moisture, though measured on a single day can be expected to be an
important factor separating plant species distribution. The non­significance of frost
and non­frost disturbance in the CCA might be explained by our choice of habitats.
The study sites were often situated in areas with fairly well established vegetation.
Highly unstable non­frost conditions were rarely recorded even in newly deglaciated
areas in the forefront of glaciers. Old landscape surfaces, having long been
deglaciated and therefore showing features of high frost activity, are absent in this
study, thus unstable frost disturbance conditions were not recorded.
Despite above mentioned restrictions and a limited dataset for vascular plants we were
able to draw some conclusions about plant species distribution on Svalbard. Plant
distribution in the Arctic is limited by a variety of abiotic factors (Savile 1960). A few
of them like pH, temperature, organic soil depth, soil moisture and the biotic factor
vascular plant cover were analysed. A minority of species was found to be widely
spread and covering a wide pH amplitude. The majority of species possessed narrow
to medium amplitudes regarding pH, yet they were widespread on Svalbard.
Widespread plants in the Arctic are not necessarily widespread due to their wide
ecological amplitude but due to a widespread habitat. Plants with narrow ecological
amplitudes might even have advantages regarding niche differentiation of plant
species as has been shown by McKane et al. (2002) concerning nitrogen sources in
arctic tundra. Instead of struggling for survival plant taxa seem well adapted to their
often harsh environmental conditions in the Arctic. Alternative lifestyles of both
generalists’ and specialists’ behaviour with respect to their ecological amplitude
facilitates this adaptation.
Study species Euphrasia wettsteinii
Possible host species of E. wettsteinii were examined in this study. Bistorta vivipara,
Salix polaris and the graminoids Poa arctica and Luzula confusa were the most frequent
species. However, only Bistorta vivipara and Poa arctica were present with ≥ 50% at
both Ossian Sars and Colesdalen suggesting that Bistorta vivipara and Poa arctica are
host species of E. wettsteinii. This is be supported by previous studies on Svalbard
and the mountain Sandalsnuten at Finse (Euphrasia frigida was studied at Finse. Due
to taxonomical changes, E. wettsteinii is the new name used for the species found in
Svalbard). Studies from Svalbard proposed that Salix polaris and Bistorta vivipara
among others, are within host range of E. wettsteinii (Alsos & Lund 1999, Alsos et al.
2004). Results from Sandalsnuten at Finse revealed that Bistorta vivipara and species
from the genera Poa and Salix are likely host species of E. wettsteinii (Nyléhn &
Totland 1999). Plots with E. wettsteinii were similar in species composition to those
plots without E. wettsteinii. Thus E. wettsteinii may have the potential to spread to a
wider area with use of different host species.
50

The temperature measurements at 10 cm soil depth in the plots with E. wettsteinii
showed a higher temperature (~0.5­1 °C) in comparison to the plots without E.
wettsteinii, however, the t­test showed no significance.
One patch of E. wettsteinii was found at Colesdalen in 1998 and nine patches were
discovered in 2002 (Alsos & Lund 1999, Alsos et al. 2004). This increase in patches
correlates with the temperature diagram which shows a peak in 1998 and one in 2002
due to increased summer temperature. By July 2007 the nine patches had spread to
one big continuous patch. Is the spread of E. wettsteinii in Colesdalen a consequence
of increased mean summer temperature on Svalbard? It is worth noting on the
mountain Sandalsnuten at Finse in Norway, growth and seed production of E.
wettsteinii were largely temperature dependent (Nyléhn & Totland 1999). However,
the effect of temperature on the spread of E. wettsteinii must be indirect due to
increased growth and quality of the host plant caused by elevated temperature. As a
consequence, the population of E. wettsteinii could actually decrease due to increased
competition if we assume that a warmer climate leads to a denser vegetation cover.
However, temperatures on Svalbard may have to rise substantially before this process
will take effect. Increased temperatures and changes in species composition can lead
to disappearance of some of the host species of E. wettsteinii and introduction of new
species. Due to the broad host range of E. wettsteinii one can assume that this might
not influence the ability to have proper species in the surroundings to parasitise
(Nyléhn & Totland 1999, Seel & Press 1993).
Dense patches of E. wettsteinii in Colesdalen were observed near bare ground mainly
caused by frost disturbance. Open soil patches are needed for seed establishment
(Molau 1993, Seel & Press 1993). Long­term consequences of climatic changes may
lead to less frost disturbance and, consequently, fewer open soil patches. The resulting
dense vegetation cover may limit the growth of E. wettsteinii in some places.
The thermophilic annual E. wettsteinii has not only the potential to spread when mean
July temperatures are rising by 1 degree Kelvin, this being the temperature increase
being observed for the last ten years, but is already expanding its habitat.
Conclusion
Species on Svalbard and Greenland possess the same ecological amplitudes,
regardless of their occurrence in the Low or High Arctic. Widespread species on
Svalbard do not necessarily possess wide ecological amplitudes and widespread plants
do not naturally display wide ecological amplitudes. These conclusions may be of
major interest when considering species responses to climate change. A variety of
arctic species may have ecological amplitudes large enough to accommodate for
climate change. A population of the thermophilic annual Euphrasia wettsteinii seems
to be spreading, perhaps due to rising temperatures on Svalbard, thereby indicating
early signs of climate change in the Arctic.
References
ACIA. 2006: Arctic Climate Impact Assessment: Scientific Report Cambridge.
51

Alsos, I. G. 2007: Frequent Long­Distance Plant Colonization in the Changing Arctic.
Science, 316.
Alsos, I. G. personal communication: Euphrasia wettsteinii.
Alsos, I. G. & Lund, L. 1999: Fjelløyentrøst Euphrasia frigida funnet i Colesdalen,
Svalbard. Blyttia, 57: 36.
Alsos, I. G., Westergaard, K., Lund, L. & Sandbakk, B. E. 2004: Floraen i Colesdalen,
Svalbard. Blyttia, 62: 142­150.
Arnesen, G. personal communication: Substrate map of Svalbard.
Bakken, T., Kålås, J. & Viken, Å. 2006: Norsk Rødliste 2006.
Billings, W. D. & Mooney, H. A. 1968: The ecology of arctic and alpine plants.
Biological Reviews, 43: 481­529.
Brochmann, C. & Steen, S. W. 1999: Sex and genes in the flora of Svalbard—
implications for conservation biology and climate change. Matematisk
Naturvitenskapelig Klasse, Skrifter, 38: 33­72.
Bråthen, K. A. & Hagberg, O. 2004: More efficient estimation of plant biomass.
Journal of Vegetation Science, 15: 653­660.
Böcher, T. W. 1963: Phytogeography of middle west Greenland. Meddelelser om
Grønland, 148.
Canoco for Windows 4.5. 1999
Coudun, C. & Gegout, J. C. 2005: Ecological behaviour of herbaceous forest species
along a pH gradient: a comparison between oceanic and semicontinental regions in
northern France. Global Ecology and Biogeography, 14: 263­270.
Ellenberg, H., Weber, H. E., Düll, R., Wirth, V., Werner, W. & Paulissen, D. (red.).
1991: Zeigerwerte der Pflanzen in Mitteleuropa. Scripta Geobotanica 18 (Lehrstul
für Geobotanik der Universität Göttingen). Göttingen.
Elvebakk, A. 1999: Bioclimatic delimitation and subdivision of the Arctic.
Matematisk Naturvitenskaplig Klasse, skrifter, 38: 81­112.
Elvebakk, A. 2005: A vegetation map of Svalbard on the scale 1:3.5 mill.
Phytocoenologia, 35: 951­967.
Gelting, P. 1934: Studies on the vascular plants of east Greenland between Franz
Joseph Fjord and Dove Bay (Lat. 73° 15’ – 76° 20’ N.). Meddelelser om Grønland,
101.
Hagen, A. R. 2001: Trans­atlantic dispersal and phylogeography of Cerastium
arcticum (Caryophyllaceae) inferred from RAPD and SCAR markers. American
Journal of Botany, 88: 103­112.
Hill, M., Roy, D., Mountford, J. & Bunce, R. 2000: Extending Ellenberg's Indicator
Values to a New Area: An Algorithmic Approach. The Journal of Applied Ecology,
37: 3­15.
52

Hultén, E. & Fries, M. 1986: Atlas of North European Vascular Plants North of the
Tropic of Cancer. Königstein, Koeltz Scientific Books.
Jónsdóttir, I. S. 2005: Terrestrial ecosystems on Svalbard: heterogeneity, complexity
and fragility from an arctic island perspective. Biology and Environmen:
proceedings of the Royal Irish Academy, 105B: 155­165.
Karlsen, S. R. & Elvebakk, A. 2003: A method using indicator plants to map local
climatic variation in the Kangerlussuaq/Scoresby Sund area, East Greenland.
Journal of Biogeography, 30: 1469­1491.
Lid, J., Lid, D. T. & Elven, R. 2005: Norsk flora, Det Norske Samlaget. 1014 s.
McKane, R., Johnson, L., Shaver, G. & Nadelhoffer, K. 2002: Resource­based niches
provide a basis for plant species diversity and dominance in arctic tundra. Nature,
415: 68­71.
MET. 2007: Døgnverdier Svalbard Lufthavn, Metorologisk institutt. Lokalisert 26.07.
2007 på World Wide Web:
http://met.no/observasjoner/svalbard/SvalbardLufthavn/index.html?stasjon.
Molau, U. 1993: Phenology and reproductive ecology in six subalpine species of
Rhinanthoideae (Scrophulariaceae). Opera Bot., 121: 7­17.
Nyléhn, J. & Totland, Ø. 1999: Effects of temperature and natural disturbance on
growth, reproduction and population density in the alpine annual hemiparasite
Euphrasia frigida. Arctic, Antarctic and Alpine Research, 31: 259­263.
PAF. 2007: Checklist of the Panarctic Flora (PAF). Draft version. I: Elven, R.,
Murray, D. F., Razzhivin, V. & Yurtsev, B. A. (red.). Oslo, Univ. of Oslo.
Przybylak, R. 2007: Recent air temperature changes in the arctic. Annals of
Glaciology, 46: 316­324.
Raup, H. M. 1965: The flowering plants and ferns of the Mesters Vig district,
northeast Greenland. Meddelelser om Grønland, 166.
Raup, H. M. 1969: The relation of the vascular flora to some factors of site in the
Mester Vig district, northeast Greenland. Meddelelser om Grønland, 176: 1­80.
Rønning, O. I. 1961: Some new contributions to the flora of Svalbard. Norsk
Polarinstitutt Skrifter, 124: 1­20.
Saetersdal, M. & Birks, H. J. 1997: A comparative ecological study of Norwegian
mountain plants in relation to possible future climatic change. Journal of
Biogeography, 24: 127­152.
Savile, D. B. O. 1960: Limitations of the competitive exclusion principle. Science,
132: 1761.
Savile, D. B. O. 1972: Arctic Adaptations in Plants. Småtrykk, 20.
Seel, W. E. & Press, M. C. 1993: Influence of the host on three sub­Arctic annual
facultative root hemiparasites. New Phytologist, 125: 131­138.
Sørensen, T. 1933: The vascular plants of East Greenland from 71° 00’ to 73° 30’ N.
Meddelelser om Grønland. 101.
53

Walker, M., Wahren, C., Hollister, R., Henry, G., Ahlquist, L., Alatalo, J., Bret­
Harte, M., Calef, M., Callaghan, T., Carroll, A., Epstein, H., Jonsdottir, I., Klein, J.,
Magnusson, B., Molau, U., Oberbauer, S., Rewa, S., Robinson, C., Shaver, G.,
Suding, K., Thompson, C., Tolvanen, A., Totland, O., Turner, P., Tweedie, C.,
Webber, P. & Wookey, P. 2006: From The Cover: Plant community responses to
experimental warming across the tundra biome. Proceedings of the National
Academy of Sciences of the United States of America, 103: 1342­1346.
Washburn, A. L. 1969: Weathering, frost action, and patterned ground in the Mesters
Vig district, northeast Greenland. Denmark. Meddelelser om Grönland, 176:1­303.
Washburn, A. L. 1973: Periglacial processes and environment. New York, St.
Martin's Press. 320 s.
54

Recruitment along retreating glaciers
Unni Vik 1 Ingelinn Aarnes 2 Eike Müller 3
1 University of Oslo, Department of Biology, Program for Molecular Ecology and Biosystematics, 0316
Oslo,NO, 2 University of Bergen, Department of Biology, Ecological and Environmental Change
Research Group, 5020 Bergen, NO 3 The University Centre in Svalbard, Arctic Biology, 9171
Longyearbyen, NO
Abstract
Glacier forelands are excellent models for studying primary successions but relatively
few have been investigated in the High Arctic. This study investigates differences in
the succession patterns in three bioclimatic zones on Svalbard. The species occurrence
was recorded on moraines of different ages at Nordre Franklinbreen, Mayerbreen, and
Kongsbreen situated in bioclimatic zone A (Arctic polar desert), B (Northern arctic
tundra) and C (Middle arctic tundra). Species composition was sampled using three to
six plots (50 x 50 cm) on each moraine as well as by compiling a total species list.
The glacier forelands in bioclimatic zone B and C follow the expected succession
pattern of species substitution with time. In bioclimatic zone A no clear succession
pattern could be found. Species richness increases in bioclimatic zone B and C with
time whilst in bioclimatic zone A it is comparatively constant and several species
occur irrespective of moraines’ age. Additional studies of glacier foreland in
bioclimatic zone A are needed to confirm the generality of this pattern.
Introduction
Arctic areas are expected to experience major changes due to global warming in the
coming years (ACIA 2005). Apart from the major climate alternations (ice ages),
arctic plants must have experienced rapid climate oscillations (Brochmann & Steen
1999) as they occurred more frequent than expected both within colder and warmer
stages (Taylor et al. 1993). Strategies of arctic plants to survive and maintain genetic
diversity during rapid changes are predominantly those involving modes of
reproduction (sexual or clonal) and development of high ploidy levels.
Plants have individualistic responses to this expected change (Erschbamer 2007). A
study in the Alps revealed that pioneer species are particularly vulnerable to global
warming (Erschbamer 2007). Walker et al. (2006) tested the responses of tundra
species to experimental warming and projected a possible loss of as much as 40% of
the current tundra vegetation by 2100. It is likely that freshly deglaciated terrain will
provide pioneer species with a sanctuary during climate change, and conserve parts of
the biodiversity in the Arctic (Moreau, submitted). If this is the case, then it is
probable that arctic species which are unable to colonize moraines are more
vulnerable to climate changes.
Polyploidy represents a mechanism of adaptation and speciation (Stebbins 1971,
Lumaret 1988). Estimated numbers of polyploid angiosperms are between 47% and
70% (Grant 1981). The proportion of plant taxa with a higher ploidy level
(polyploids) is in general showing an increasing trend at higher latitudes in the
55

northern hemisphere (Brochmann et al. 2004). In addition, the proportion of
polyploids seems to be particularly high in previously glaciated areas. Out of the 161
species indigenous to Svalbard, 78% are polyploids (Brochmann & Steen 1999). One
reason for this high number is that polyploidy secures the maintenance of genetic
diversity in marginal and climatically unstable environments (Brochmann & Steen
1999). Therefore, it can be expected that polyploid species occur more frequently on
newly deglaciated terrain than diploid species.
In harsh environments, such as the Arctic, with high variability in space and time,
particularly in the length of the growing season, sexual reproduction appears to be
selected for (Bliss & Gold 1999). However, it has been found that seedling mortality
is very high on glacier forelands (Niederfriniger et al. 2000), and thus it is possible
that there is a difference in the extent of sexual reproduction amongst species that
specialize in colonizing freshly deglaciated terrain compared to other areas in the
High Arctic.
The most significant characteristic of the Arctic is the absence of trees for climatic
reasons (Körner 2003). In the High Arctic (bioclimatic zone A, B and C) even erect
shrubs are absent. Thus, the succession in the High Arctic differs significantly from
vegetation development in tree­ and shrub­dominated areas. Vegetations without trees
and erect shrubs can be influenced and altered by factors more easily and have
different pathways for succession depending on the factors altering the development.
The use of glacier forelands to study succession by substituting space for time is a
well established approach (Matthews 1992). Primary succession in the Arctic follows
patterns that are also seen in glacier forefields in Norway and the Alps (Matthews
1992). Species richness increases with time in Svalbard, despite a loss of mid
successional species (Hodkinson et al. 2003). Oppositely, studies in the central Alps
found that species richness reaches a peak at 40­50 years and declines towards the
climax (Raff et al. 2006). This early peak in species richness has been observed in
studies conducted in Norway, the Alps (Matthews 1992) and in the Canadian Arctic
(bioclimatic zone C) (Jones & Henry 2003). This fall in richness after the early peak
is often attributed to competition in areas where environmental conditions are
favourable and ground cover is close to 100%, whilst in more severe conditions an
absence of a distinct peak can reflect that plants at all successional stages experience
rather uniform conditions (Matthews 1992).
Observations of succession on glacial forelands on Svalbard in bioclimatic zone B and
C have shown that there is a clear relationship between species changes and time
(Moreau, submitted). However, Jones and Henry (2003) concluded that succession
may not follow the expected directional replacement of species with time in more
extreme polar­desert environments. There have been few studies in the High Arctic
and so far no studies have been conducted in bioclimatic zone A.
This study used glacial forelands to record primary succession patterns in three
bioclimatic zones on Svalbard investigated whether or not species richness increases
with time in the High Arctic. Is there is a difference in the extent of sexual
reproduction and polyploidy among species, occurring on freshly deglaciated terrain
than elsewhere on Svalbard?
56

Figure 1. Location of the three glacier forelands and a site rich in species.
Sites
Svalbard is highly influenced by warm north Atlantic currents and is warmer than
other areas at such high latitudes. The Archipelago covers several bioclimatic zones
(Elvebakk 1999) and has a varied climate: The mean (1961­1990) July temperature
varies between 1.9° – 6.5° C and the mean annual precipitation is between 190­490
mm (Norsk meterologisk institutt). There are 164 vascular species present at Svalbard
(Rønning 1996) (161 according to Brochmann & Steen 1999).
Three glacier forelands (Fig. 1) were studied. The three bioclimatic regions are very
different in both climate and vegetation (see Tab.1 and 2). Nordre Franklinbreen in
Franklinfjorden at Nordaustlandet is situated in bioclimatic zone A. With the
exception of the youngest moraine it was difficult to distinguish individual moraine
ridges. Four sites were picked in this area; only the youngest moraine was a clearly
defined ridge. The main vegetation in the region is Luzula confusa polar desert
(Elvebakk 2005).
In zone B, Mayerbreen in Krossfjorden, two lateral moraines were studied. The young
moraine has gentle slopes and consists of bedrock with a relatively thin morainal
cover. The analysis was conducted on stable summits. The older moraine extends to
approximately 100 meters above sea level and has a steep north­west facing slope.
The main vegetation in the region is mesic acidic Luzula confusa tundra (Elvebakk
2005)
At Kongsbreen in Kongsfjorden three moraines were studied. A very recently
deglaciated moraine, an intermediate­age moraine, that represents the glacier front in
1869 (Lamont 1876 in Lefauconnier 1991,) and an old moraine. These occur within
57

an enclave of bioclimatic zone C within zone B. Cassiope tetragona tundra is the main
vegetation type (Elvebakk 2005).
Colesdalen in Isfjorden is not a glacier foreland area; it is situated in bioclimatic zone
C and is considered among the most species rich sites in all of Svalbard. The site was
included to have a diversity reference for well established, rich arctic tundra
vegetation. One south­facing slope dominated by Cassiope tetragona was analysed.
For appropriate comparison 9 out of 22 plots were used.
Table 1. Characteristics of study sites including main vegetation types.
Vegetation Type
Estimated
distance to Bioclimatic
Site Slope of Moraine
(Elvebakk, 2005)
Glacier* Zone Date
Nordre
Franklinbreen
Northwest Luzula confusa Polar Desert 300 m ­ 1200 m A 20.07.2007
Mayerbreen Southwest Luzula confusa Mesic acidic Tundra 300 m ­ 650 m B 17.07.2007
Kongsbreen South / southwest Cassiope tetragona Tundra 800 m ­ 2500 m C 21.07.2007
*Distances estimated using ArcGIS, lowest number represents the distance from the youngest morain
to tha glacier, highest the distance to plots in established vegetation.
Table 2. Overview of all sites studied in the survey and coordinates of sample plot areas.
Site Abbreviation Location (UTM)
Altitude
(masl) Age # Plots
Nordre
Franklinbreen
N1 33x 0588861/8901802 53 Younger moraine 3
Nordre
Franklinbreen
N2 33x 0588725/8901772 57 Older moraine 3
Nordre
Franklinbreen
N3** 33x 0588282/8901964 23
Established vegetation with
nutrient input
­
Nordre
Franklinbreen
N4 33x 0587985/8901644 12 Established vegetation 3
Mayerbreen M1 33x 0440859/8800766 25 Younger moraine 5
Mayerbreen M2* 33x 0440870/8800608 87 Older moraine 3
Mayerbreen M3* 33x 0440825/8800578 100 Older moraine 4
Kongsbreen K1 33x 0447365/8761710 22 Young moraine 3
Kongsbreen K2** 33x 0446509/8761654 47 Young moraine ­
Kongsbreen K3 33x 0445660/8762902 120 Intermediate moraine 3
Kongsbreen K4 33x 0445378/8763388 70 Established vegetation 3
*At these two sites sample plots were taken but occur in the total species list as one site, ** Recorded
at the total species list as separate site but no plots analysed
58

Methods
The vegetation was analysed with the intercept point frame analysis (Bråthen &
Hageberg 2004). At each site a number between three and six plots were analyzed
using a 50 x 50 cm steel frame with a grid of 10 cm. Vascular plants were recorded at
25 cross sections of the grid, x. The plots were distributed randomly within a specified
homogenous vegetation unit. The overall vegetation cover including vascular plants,
bryophytes, lichens, cryptogram crust and bare ground was estimated by eye on a ten
percent step scale. Additionally, all species present at each site were recorded. Plant
communities were visually evaluated and all plant species growing within the
homogenous communities were listed. The size of the homogenous areas varied
depending on plant composition in the communities. Small scale changes in plant
communities resulting from microhabitats produce minor dissimilarities, which were
accepted within homogenous areas.
To obtain an approximate chronological control of the moraines, aerial photos, maps
with dated teminal positions and lichenometry were used. Lichenometry is based on
the assumption that the growth rate of crustose lichen species (most often
Rhizocarpon rhizocarpon) has a linear relationship with time (Matthews 1992), i.e.
larger thalli are older than smaller ones. Growth curves have been constructed for
many areas as the growth pattern varies in different areas (Bradwell and Armstrong
2007). Growth curves are made by measuring lichens growing on substrates of known
ages, for example gravestones (Matthews 1992). If crustose lichen is found on the
surface being studied, an age estimate can be made. In this study the mean of the five
largest thalli of Rhizocarpon found on each site were compared to the local lichen
growth curve for North­West Spitsbergen (Werner 1991).
To compare plant diversity between sites the alpha (i.e. species richness) diversity of
vascular plants as described in Whittaker (1972) was calculated using presents/
absence data. Additionally, the Shannon’s diversity index H and the species evenness
(Shannon’s evenness index also known as Pielou's evenness index) were computed
(see Tab. 3). For the derivation of Shannon’s index frequency data of the intercept
point frame analysis and the program PAST was applied (Hammer & Harper 2003).
The different numbers of plots recorded at each site resulted in an unbalanced design,
for data analysis only data of the first three recorded plots at each site were used. With
a small number of plots recorded at each site, plot data were applied only for
calculating α diversity, Shannon’s H and E indices.
Observed frequencies of polyploid and diploid species as well as clonal and sexual
reproduction were compared to the data from Svalbard (Brochmann & Steen 1999)
using the G­test for reasonable size (see Tab. 3). For analysing diploid and polyploid
ratios as well as sexual and clonal ratios, the total species list of each site has been
used. The property of each species (polyploidy, reproduction mode) was derived from
Brochmann & Steen’s (1999). Data for comparison were derived from Hodkinson et
al. (2003).
Spearmans­s rank­order correlation was chosen to analyse the relationships between
plant succession stage, ploidy, and reproduction mode. Since this is a test requiring an
approximately continuous scale, data were tested applying the Kolmogorv­Smirnov
Test. Calculations and tests were made using the software package SPSS 10.0.1.
59

Table 3. Formulas of the different applied indices (n­number of individuals, S­number of
species, N­total numbers of individuals and p­relative abundance of each species) and
G­test (Oi is the frequency observed in a cell, E is the frequency expected on the null
hypothesis)
Test/ Index Formula
Shannon Index H H ' = − ∑
Species Evenness E
G­Test
60
S
i = 1
pi ln pi
H '
E =
H ' max
∑
G =
2 Oi o ln( Oi / Ei )
i

Results
Succession
Nordre Franklinfjorden – bioclimatic zone A
Cerastium regelii and Saxifraga oppositifolia are found only on the earliest moraine.
No species grow exclusively on the intermediate stage surface. Saxifraga foliolosa
and S. rivularis are present in both intermediate and mature vegetation assemblages.
Cardamine bellidifolia and Luzula confusa are found on all moraines, whilst Papaver
dahlianum as well as Stellaria longipes appear on three of four sites surveyed.
Mayerbreen – bioclimatic zone B
The very early colonizers on the young moraine but not present on the intermediate
stage surface are Draba alpina, D. nivalis, Phippsia algida, Saxifraga oppositfolia,
and Trisetum spicatum. Many species are present on both the early moraine and in the
intermediate moraine such as Cochlearia groenlandica, Draba corymbosa, Luzula
confusa, Oxyria digyna, Poa alpina, Poa arctica, Saxifraga cernua, S. cespitosa, S.
hyperborea, S. nivalis, S. rivularis and Stellaria longipes. Only a few species are
exclusively present on the intermediate vegetation, such as Draba lactea, D.
oxycarpa, and Papaver dahlianum. Noteworthy absences in the intermediate stage are
the otherwise ubiquitous Saxifraga oppostitifolia and Salix polaris, which were not
recorded at Mayerbreen at all.
Due to the instability of the steeply sloped intermediate/older moraine the vegetation
was very patchy, which consisted of gramnoids and pioneers.
Kongsbreen – bioclimatic zone C
At Kongsbreen the early colonizers were Braya purpurascens, Draba nivalis, D.
subcapitata, Papaver dahlianum, Sagina nivalis, Saxifraga aizoides, S. cespitosa, S.
oppositifolia and Silene uralensis. Only Draba spp., Sagina nivalis and Silene
uralensis grow exclusively on the young moraines, while Braya purpurascens,
Cerastium arcticum, and Saxifraga cespitosa persist in the intermediate stages.
Several species have been found only on intermediate moraines (Cochlearia
groenlandica, Draba arctica, D. corymbosa, Silene furcata and Saxifraga nivalis) or
mature vegetation (Arenaria pseudofrigida, Bistorta vivipara, Carex nardina,
Cassiope tetragona, Dryas octopetala, Luzula nivalis, Silene acaulis and Trisetum
spicatum). Some species are ubiquitous at all successional stages and present on
moraines of all ages. Saxifraga oppositifolia and Salix polaris are present on the
earliest moraine and persist, albeit less abundantly, in the climax vegetation of Dryas­
Cassiope heath.
Diversity
To get a general impression of the increase in species from younger to older stages of
the succession at all moraines, the species numbers (alpha diversity) of the different
bioclimatic zones are compared (Fig. 2a, b). It is evident (Fig. 2a) that the number of
plots analysed were not sufficient as only the curve for cumulative species numbers
for Nordre Franklinbreen flattens off. A non­flatted curve indicates that the whole
range of diversity were not present in the plots that were sampled at Mayerbreen and
Kongsbreen.
61

Figure 2. a: Cumulative number of species observed with the point intercept frame surveys,
every dot is a succeeding plot. b: alpha diversity of different succession stadiums at
three glacier forefields.
In early pioneer stages the species numbers are low to zero (Fig. 2b). None or a very
few vascular plants and few scattered bryophytes were observed at the very recently
deglaciated moraines. At these young moraines small streams seem to be important
for starting plant establishment. They offer numerous microhabitats that facilitates the
establishment of pioneer plants. Intermediate and older stages are characterised by an
increase in species numbers and greater plant cover. The cumulative number of
species per plot increased continuously except on an older moraine at Nordre
Franklinsbreen (fig 2a). In bioclimatic zone B and C (Mayerbreen, Kongsbreen)
species numbers increase very little from intermediate to old stages (Fig. 2b). This
pattern was not found in bioclimatic zone A, where species showed the greatest
increase from intermediate to old moraines. Additionally, all stages in zone A are very
similar in their species composition.
Besides the increase in species from younger to older stages, a highly significant
correlation between the moraine’s age and plant cover was observed (Fig. 3a).
62

Figure 3. a: Total plant cover against estimated age of the moraines Spearman’s rank order
correlation coefficient 0.758, significant at the 0.001 level, Rsq=0,79 b: Shannon index
and species evenness for Colesdalen, Mayerbreen, Nordre Franklinbreen and
Kongsbreen
Although Mayerbreen is situated in the bioclimatic zone B, the moraines have a
higher alpha diversity than other moraines in this study. The species evenness E and
the Shannon index H differs little from Colesdalen, one of the richest places in flora
on Svalbard situated in bioclimatic zone C (fig 3b). The lowest index as well as the
lowest evenness was calculated for Nordre Franklinsbreen which reflects the
dominance of Luzula confusa and the low number of species and their scattered
distribution.
The glacier forefields of the colder arctic zones (A, B) favoured generalist species
(Fig. 4). In bioclimatic zone A early colonizers are still present in later succession
stages, whereas in bioclimatic zone C early colonizers are almost absent from well
developed, closed communities. The high proportion of species present at both
younger and older sites in bioclimatic zone B indicates a high similarity or a failure to
visually classify the age of the moraines and successional stage correctly.
63

Figure 4. Presence and absence of all species observed in this study on glacier forefields. Species abbreviations consist of first the three genus and first three
species letters
64

Ploidy
The frequencies of diploid/ polyploid species occurring on each moraine have been
plotted against the estimated age of the moraine (Fig 5a). The age of the moraines and
the number of polyploidy species are not significantly correlated. Nevertheless, the
data indicate a slight trend to a lower frequency of polyploids on older moraines.
Hodkinson et al. (2003) found a highly significant correlation (Tab. 4) for a lower
level of polyploid species in older succession stages at glacier forefields (Fig 5b). A
slight non­significant trend towards a larger number of polyploids at glacier forefields
was found from bioclimatic zone C to A (data not shown).
Figure 5. a: Part of polyploid plants in percent against the estimated age of the moraines own
survey Rsq=0.16 b: part of polyploidy plants against radiocarbon dated moraines
derived from Hodkinson et al. (2003) Rsq=0.88.
A G­test of the diploid/ polyploid frequencies between frequencies on moraines
observed in this study and published frequencies from all plants present in Svalbard
(Brochmann & Steen 1999) revealed that the frequencies observed from the authors
and frequencies from all plants in Svalbard show no differences (p value 0.15; error
probability 0.05). Exception is one observed frequency (Kongsbreen climax stage)
which has been compared to the frequency of all plants in Svalbard; this frequency
differs significant from the frequency of all plants in Svalbard (p value 0.012)
Table 4. Correlations between estimated age of moraines ground cover, ploidylevel, and sexual
reproduction. AGE_EST­estimated age, ploidy%­percentage of polyploid plants,
sexual%­percentage of sexual reproducing plants. First row own data, second row data
Hodkinson et al. (2003).
ploidy % sexual %
Age estimation Correlation Coefficient 226 ­630
Significance. (2­tailed) 530 51
Number 10 10
Age estimation Correlation Coefficient 0,919** ­,986**
Significance. (2­tailed) 3 0
Number 7 6
** Correlation is significant at the 0.01 level
Reproduction
No significant correlation was found between the age of the moraines and the mode of
reproduction (Fig. 6a) but the frequency of sexually reproducing plants is higher in
younger moraines. These findings are supported by the data of Hodkinson et al.
65

(2003), from which a highly significant correlation between proportion of sexual
reproducing plants and the age of the succession has been calculated (see Tab. 4). A
higher proportion of plant species on younger moraines reproduce sexually than on
older terrain, where the number of clonally reproducing species increases (Fig 6b).
Figure 6. a: Plot of sexual reproducer against the estimated age of the moraines own survey
Rsq=0.33 b: plot of sexual reproducer against radiocarbon dated moraines Hodkinson
et al. (2003) Rsq=0.95
The frequency of species mainly reproducing sexually tended to be higher in warmer
zones than in colder zones (data not shown). A G­test demonstrates that the overall
observed frequency for plants on glacier forefields is not the same as for all plants
found in Svalbard (null hypothesis rejected). A smaller proportion of plants growing
in glacier forefields reproduce clonally than all plants known to Svalbard.
Discussion
Succession and diversity
The observed vegetation pattern and biodiversity level at the different moraines in
bioclimatic zone B and C are generally congruent to earlier studies in high arctic areas
(Moreau submitted, Moreau 2005, Hodkinson 2003, Jones & Henry 2003). At all
bioclimatic zones (A­C) in the early stages Saxifraga spp. are frequent. The members
of the Brassicacea family, in particular the genera Draba and Cardamine, are less
frequent but a constant part of the pioneer vegetation. Early pioneers were found to be
the same as postulated by the mentioned authors (see results Fig. 5). Succession stages
in zone C and B clearly differ in species composition with time, whilst zone A did not
show the expected species substitution with time pattern.
At both Kongsfjorden and Mayerbreen a species substitution succession occurs. This
follows the general pattern expected in glacier forelands both in Europe (Matthews
1992), and Svalbard (Moreau submitted, Moreau 2005 and Hodkinson et al. 2003).
There are typical early colonizers like the widespread species Saxifraga oppositifolia
as well as species like Sagina nivalis and Draba spp. (Fig. 5). These species specialize
in colonising fresh moraines but are unable to persist in the intermediate and mature
vegetation. It is difficult to estimate for certain why these species are unable to persist
in the intermediate stage moraines as these are still open communities and have
66

patchy vegetation so competition is not likely to be a factor. Hodkinson et al. (2003)
found that there were changes over time with increasing soil organic matter, nitrogen
and soil moisture as well as a decreasing soil pH. The niche of these species can be
related to all of these environmental variables.
Several species are typical at intermediate stage moraines, like Papaver dahlianum
and Saxifraga spp., which do not appear in the more mature vegetation stage. This
species substitution/ reduction can be related to the species demand for open areas or
interspecies competition, as in bioclimatic zone B and C in more developed stages the
vegetation cover gets more dens. Similar, few early colonizers have been found within
climax (e.g. Dryas octopetala and Cassiope tetragona tundra communities),
indicating that they are poor competitors. In the stable vegetation of the Dryas –
Cassiope heath communities on old moraines at Kongsbreen, particularly, Draba spp.
were not observed, which suggests that in stable climax vegetation the pioneer species
might be out competed. The species that only appear at a single stage may, to lesser or
greater extent, represent specialist of particular habitats, although species like
Cassiope tetragona are very widespread in Svalbard (www.svalbardflora.net).
Some species are ubiquitous, appear in all succession stages and are present on
moraines of all ages. Saxifraga oppositifolia and Salix polaris occur on the earliest
moraines and persist, less dominant, in the climax vegetation. S. polaris and S.
oppositifolia as well as Luzula confusa (only in zone A and B) are seen as widespread
(Moreau submitted, Hodkinson et al. 2003).
At Mayerbreen the absence of the widespread S. oppostitifolia, and S. polaris is
noteworthy and cannot be explained with its steepness. Since both species are among
the earliest colonizers on moraines and thus must have a high degree of tolerance for
disturbance. It may be related to other reasons as soil or a simple observation bias or
to chance events. The steepness of the moraine is the reason for the high species
richness. The patchy slope caused by erosion seems to create many different micro
habitats. Although situated in bioclimatic zone B the Shannon’s diversity index and
the species evenness are similar to Colesdalen, with a rich flora in bioclimatic zone C
(Alsos et al. 2004).
At the Nordre Franklinbreen the succession development is very different from the
general pattern seen in Kongsfjorden, Mayerbreen and also in the studies conducted at
other glacier forelands in Svalbard (Moreau submitted, Moreau 2005 and Hodkinson
et al. 2003). There is no clear species succession with time visible in the record, and
many species appear in all stages seemingly independent of the time factor. Polar
deserts and semideserts are marginal for establishment of vascular plants (Svoboda &
Henry 1987), a major landscape feature of bioclimatic zone A. Obvious in the glacier
foreland of Nordre Franklinsbreen is that the plants are scattered even at older
moraines and the vegetation consists of small patches.
One mature site (N3) is clearly set apart in the Norde Franklinbreen foreland. Many
species are present here that are absent elsewhere on this foreland However, this site
is influenced by higher nutrient availability as there are small ponds with several eider
ducks (Somateria mollissima) as well as reindeer (Rangifer tarandus platyrhynchus)
judging from the droppings present.
67

The intermediate stages seem to have many specialized species in all sites with the
exception of Nordre Franklinbreen where no particular species are on this stage. There
is no overlap in the species that occur in the intermediate stage at both Mayerbreen
and Kongsfjorden. Many species occur at intermediate stage in Kongsfjorden and at
both stages in the Mayerbreen foreland. It is possible that this is related to
Mayerbreen’s intermediate stage moraine being very steep and unstable. However,
this can be evaluated as unlikely because most species present at more or less newly
deglaciated terrain are generally adapted to a high disturbance regime. Another
possible explanation is that the early moraine at Mayerbreen may be slighlty older
than the early moraine in Kongsfjorden so that a direct comparison between them is
complicated.
As indicated in the results chapter in the development of the plant communities (see
Fig. 4) in bioclimatic zones A and B a higher number of widespread species occur in
both zones. And thus the glacier forelands in the colder arctic zones (A and B) are
thought to favour generalist species.
Polyploidy
Hybridization and polyploidy generate new races and species, some which become
adapted to the new conditions in regions vacated by ice (Brochmann et al. 2004). The
frequency of polyploids increases with latitude in the northern hemisphere and the
proportion of polyploids is particulary high in previous glaciated areas (Brochmann &
Steen 1999). In the Arctic polyploidy secures stable reproduction and maintenance of
genetic diversity in marginal and climatically unstable environments (Brochmann &
Steen 1999). Our study spans from bioclimatic zone C in south (Kongsbreen) up to
the northern bioclimatic zone A at Nordre Franklinbreen and includes moraines of
different ages. The data obtained during the study did not give any significant
correlations on the polyploidy proportion increasing with neither bioclimatic zone nor
with the age of moraines. However, they show a trend towards an increasing number
of polyploid species with latitude, in accordance to studies of Brochmann et al.
(2004), Brochmann & Steen (1999) and Stebbins (1984). The high significant
correlations calculated from data derived from Hodkinson at al. (2003) supports the
trends mentioned above and also the hypothesis of Stebbins (1985) that polyploids are
more successful than diploids in colonizing terrain after deglaciation.
Additionally, the result of the G­test supports the above mentioned hypothesis. In
particular the observed frequencies at the climax vegetation of Kongsbreen, which
show a significantly different poly/ diploid ratio than glacier forefield’s pioneer
vegetation, confirm that polyploidy is the strategy of choice in such environments.
Reproduction
Most species found in Svalbard have mixed reproductive systems with varying
frequencies of asexuality, sexual reproduction via self­pollination, and “normal”
sexuality via cross­pollination (Brochmann & Steen 1999), although 97 of the 161
species are mainly sexual reproducers (Brochmann & Steen 1999). However, a G­test
revealed that the observed plants in glacier forefields have not the same frequencies of
reproduction modes as the sum of all plants found in Svalbard. In addition, the trend
towards higher sexual reproduction in younger succession, which is confirmed by the
correlation calculated from Hodkinson’s data (Hodkinson et al. 2003), suggests that a
sexual reproduction strategy is advantageous for plant establishment on moraines. The
68

frequency of sexual reproduction of plants at glacier forfields appears to be higher
than the Svalbard average. It is likely that the number of species reproducing sexual
on glacier forefields is higher because seeds are able to survive harder conditions,
have a greater advantage in establishment and dispersal.
The observed trend towards lower frequencies of sexual reproduction in colder
climates can be either explained by the limited resources, which have to be allocated
for sexual reproduction or shortcomings in this reproduction strategy (lack of
pollinators, establishment rate of seeds extremely low, lack of mature seeds because
of season length or low temperature).
Climate change
How plant communities and succession will react to climate change is difficult to
predict, as there is a highly complex interaction between species ecological amplitude
and competition. The only possibility is to formulate assumptions about presence and
absence in future plant communities. However, it is important to postulate which
species are most vulnerable to climate change in high arctic areas. The chance is given
that glacier forelands can become refugia for high arctic species that are highly
adapted to cold climates and as such will probably suffer because of the predicted
global warming. Therefore, the species with a northerly distribution that are rare and
in addition not able to colonize newly deglaciated terrain are likely to be the species
most vulnerable to climate change.
By using data from glacier foreland studies it is possible to distinguish which species
are limited to colonize glacier forelands. By comparing this to the species that are
found in bioclimatic zone A (Elven et al. 2003), with distribution on Svalbard
(Svalbardflora.net 2007) as well as a narrow circumpolar distribution (Hultèn 1962) it
was found that the following species could potentially be most vulnerable to climate
change in Svalbard: Alopecurus borealis, Draba norvegica, Eutrema edwardsii,
Festuca hyperborea, Minuartia rossii, Poa abbreviata ssp. abbreviate, Potentilla
pulchella ssp. pulchella, Ranunculucs nivalis and Saxifraga hirculus. However,
Alopecurus borealis, Draba norvegica, Eutrema edwvadsii, Minuartia rossii,
Potentilla pulchella, and Ranunculus nivalis are more widespread on Svalbard than
the other species mentioned. are more or less common in other areas than Svalbard
and are not considered to be particularly vulnerable on a circumpolar basis. Minuartia
rossii is on the red list of Svalbard and already a vulnerable species on the
archipelago. It has a wide distribution (ranging from bioclimatic zone A­E) and
therefore difficult to predict how it will respond to an increase in temperature. Festuca
hyperborea and Poa abbreviata have a more narrow distribution and are not
particularly thermophilic. An increase in temperature may put these species at risk, in
particular Festuca hyperborea which already is on the red list of Svalbard.
Conclusion
Glacier forelands are excellent places for studying primary succession by substituting
space for time. The glacier forefield in bioclimatic zone A did not show the expected
species substitution with time succession pattern. Instead many species occurred
irrespective of the relative ages of the moraines. Whereas species substitution during
succession follows general patterns in the arctic bioclimatic zones B and C. There is a
significant difference in the extent of sexual reproducers in this study compared to
69

Svalbard in general. Moreover polyploidy frequencies are slightly higher in northern
bioclimatic zones and on younger moraines.
References
Alsos, I. G., Westergaard, K., Lund, L. & Sandbakk B. E. 2004: Floraen i Colesdalen.
Blyttia 62, 142­150
Arctic Climate Impact Assessment 2005: Cambridge University Press
Bliss, L. C. & Gold, W. G. 1999: Vascular plant reproduction, establishment, and
growth and the effects of cryptogrammic crusts within polar desert ecosystems,
Devon Islands, N. W. T., Canada. Canadian Journal of Botany 77, 623­636
Bradwell, T., Armstrong, R. A. 2007: Growth rates of Rhizocarpon geographicum
lichens: a review with new data from Iceland. Journal of Quaternary Science 22,
311­320
Brochmann, C., Brysting, A. K., Alsos, I. G., Borgen, L,. Grundt, H. H., Scheen, A­C.
& Elven, R. 2004: Polyploidy in arctic plants. Biological Journal of the Linnean
Society 82, 521­536
Bråthen, K. A. & Hageberg O. 2004: More efficient estimations of plant biomass.
Journal of Vegetation Science 15, 653­660
Cannone, N.,Guglielmin, M. & Gerdol, R. 2004: Relationships between vegetation
patterns and periglacial landforms in northwestern Svalbard. Polar Biology 27: 562­
571
Elvebakk, A. 1999: Bioclimatic delimitation and subdivision of the Arctic. Det
Norske Videnskaps­Akademi. I. Matematisk­Naturvitenskapelig Klasse. Skrifer. Ny
serie 38, 81­112
Elvebakk, A. 2005: A vegetation map of Svalbard on the scale 1:3.5 mill.
Phytocoenologia 35, 951­967
Elven, R., Murray, D. F., Razzhivin, V. Y. & Yurtsev B.A. 2003: Checklist of the
Pan­Arctic Flora (PAF). Vascular Plants, Draft Version. University of Oslo, Oslo
Erschbamer, B. 2007: Winners and losers of climate change in a central alpine glacier
foreland. Arctic, Antarctic and Alpine Research 39, 237­244
Gough, L. 2006: Neighbor effects on germination, survival, and growth in two arctic
tundra communities. Ecography 29, 44­56
Grant V.1981: Plant Speciation. New York: Columbia Univ. Press. 2 nd ed
Grundt, H. H., Kjølner, S., Borgen, L., Rieseberg, L. H. & Brochmann, C. 2006: High
biological species diversity in the arctic flora. Proceedings of the National Academy
of Sciences of the Unites States of America 103, 972­975
Hammer, Ø., Harper, D. A. T. 2003: PAST, version 1.10
http://folk.uio.no/ohammer/past
70

Hodkinson, I. D., Coulson, S. J. & Webb, N. R. (2003) Community assembly along
proglacial chronosequences in the high Arcic: vegetation and soil development in
north­west Svalbard. Journal of Ecology 91, 651­663
Hultèn, E. 1962: The circumpolar plants. I: Vacsular cryptogams, conifers,
monocotyledons. Stockholm, Almquist & Wiksell
Jones. G. & Henry, G. H. R. 2003: Primary plant succession on recently deglaciated
terrain in the Canadian High Arctic. Journal of Biogeography 30, 277­296
Körner, C. 2003: Alpine Plant Life. Functional Plant Ecology of High Mountain
Ecosystems. Springer­Verlag, Berlin Heidelberg
Lefauconnier, B. 1991: Recent fluctuations of glaciers in Kongsfjorden, Spitsbergen,
Svalbard (79 °N). Inter­Nord 19, 449­453
Legendere, P., Borcard, D., Peres­Netro, P. R., 2005: Analyzing beta diversity:
partitioning the spatial variation of community composition data. Ecological
Monographs, 75, 435­450
Lumaret R. 1988: Adaptive strategies and ploidy levels. Acta Oecol. Oecol. Plant.
9:83­93
Matthews, J. A. 1992: The ecology of recently deglaciated terrain. A geoecological
approach to glacier forelands and primary succession. Cambridge, Cambridge
University Press
Matthews, J. A. 2005: Little Ice Age glacier variations in Jotunheimen, southern
Norway: a study in regionally controlled lichenometric dating of recessional
moraines with implications for climate and lichen growth rates. The Holocene 15, 1­
19
McCarthy, D. 1999: A biological basis for Lichenometry? Journal of Biogeography
26, 379­386
Moreau, M., Laffly, D., Joly, D. & Brossard, T. 2005: Analysis of plant colonization
on an arctic moraine since the end of the little Ice Age using remotely sensed data
and a Bayesian approach. Remote Sensing of Environment 99, 244­253
Niederfriniger, R. & Erschbamer, B. 2000: Germination and Establishment of
Seedlings on a Glacier Foreland in the Central Alps, Austria. Arctic, Antarctic and
Alpine Research 32, 270­277
Norsk Meterologisk Institutt.2007: www.met.no
Raffl,C., Mallaun, M., Mayer, R., Erschbamer, B. 2006: Vegetation Succession
Pattern and Diversity Changes in a Glacier Valley, Central Alps, Austria. Arctic,
Antarctic, and Alpine Research 38, 421­428
Stebbins, G. L. 1984: Polyploidy and the distribution of the arctic­alpine flora: new
evidence and a new approach. Botanica Helvetica 94, 1­13
Stebbins, G. L. 1971: Chromosomal Evolution in Higher Plants. London: Addison­
Wesley
Svalbardsflora. 2007: www.svalbardflora.net
71

Svoboda, J., Henry, G. H. R. 1987: Succession in marginal arctic environments.
Arctic and Alpine Research 19, No 4, 373­384
Walker, M. D., Wahren, H. C., Hollister R. D., Henry, G. H. R., Ahlquist, L. E.,
Alatalo, J. M., Bret­Harte, S. M., Calef, M. P., Callaghan, T. V., Carroll, A. B.,
Epstein, H. E., Jónsdóttir, I. S., Klein, J. A., Magnússon, B., Molau, U., Oberbauer,
S. F., Rewa S. P., Robinson, C. H., Shaver, G. R., Suding, K. N., Thompson, , C. C.,
Tolvanen, A., Totland, Ø., Turner, P. L., Tweedie, C. E., Webber, P. J. & Wookey,
P. A. 2006: Plant community responses to experimental warming across the tundra
biome. Proceedings of the National Academy of Sciences of the Unites States of
America 103, 1342­1346
Werner, A. L. 1990: Lichen growth rates for the northwest coast of Spitsbergen,
Svalbard. Arctic and Alpine Research 22, 129­140
Whittaker, R.H. 1972: Evolution and measurement of species diversity. Taxon, 21,
213­25
72

High nutrient levels can compensate for the growth­
limiting effect of low temperatures
Elke Morgner 1 and Christian E. Pettersen 2
1 Department of Physical Geography and Quaternary Geology, Stockholm University, SE­106 91
Stockholm, Sweden. 2 Department of Biology, University of Oslo, N­0316 Oslo, Norway .
Abstract
Plant growth in the Arctic is strongly influenced by low temperature and associated
short growing season, both of which also reduce soil nutrient availability. Since low
temperature and low nutrient availability may interact, it is difficult to distinguish the
impact of each of these variables. The aim of this study was to test whether
temperature per se controls plant growth or whether growth is constrained by
temperature­induced nutrient shortage. Using plant and leaf size as simple proxies for
vigor, 778 individuals of 19 different vascular plant species were examined in 10
geographical regions on Svalbard differing in mid­summer temperature and nutrient
availability. The results of this study indicate that nutrient abundance under bird cliffs
stimulates growth in all but the coldest site. We conclude that nutrient addition can
mitigate low temperature effects on growth and productivity even in the high arctic,
but when temperatures approach the thermal limit for higher plant life, temperature
itself sets the limit rather than nutrients.
Nomenclature: PAF 2007.
Introduction
The slow growth of plants in the arctic and its causes have been the subject of many
studies (Chapin 1983; Jonasson et al. 1993; Henry et al. 1986; Billings 1987). Several
of the abiotic factors that are characteristic for arctic and alpine environments are
known to restrict plant growth, but the interaction of these factors and their influences
on plant growth are complex and still subject to debate (Körner 1989). In this context
it should be mentioned that the degree of intraspecific variation in phenotypic
appearance, the phenotypic plasticity of a species, can differ between species, i.e.
species exhibit different growth responses to changes in environmental conditions
(Chapin & Shaver 1985).
Plants in the arctic are physiologically extremely well adapted to low temperatures
(Chapin 1983). As Chapin (1983) and Körner (2003) highlight, the maximum and
average relative growth rate of certain arctic and alpine tundra species is as high as in
their temperate counterparts if expressed per day of growing season. Thus,
productivity of arctic ecosystems can be as high as in any other non­water limited
natural ecosystem on earth if one accounts for the duration of the growing periods
(Körner 2003). In the extreme high Arctic, plants get increasingly smaller until they
ultimately vanish. Low nutrient levels, in particular low nitrogen availability (Henry
et al. 1986, Odasz 1994), are known to be a major limitation for plant growth in arctic
ecosystems, and it may well be that low temperature induces nutrient limitation. On
the other hand, plant nutrient absorption in cold climates was shown to be strongly
73

limited by nutrient availability rather than by low temperature (Chapin 1983).
Nitrogen availability in turn is influenced by a range of abiotic and biotic factors
which result from indirect effects of low temperatures such as slow chemical
weathering, poor soil aeration and slow organic matter decomposition (Chapin 1983).
Thus temperature acts upon plant growth in various ways (Fig. 1). The international
tundra experiment “ITEX” (Arft et al. 1999) has shown that plants from both arctic
and alpine tundra are sensitive to raised summer temperatures, and respond by
increasing vegetative growth in the short term. In the long term however, such
responses may be limited by nutrients.
Figure 1. The direct and indirect effects of temperature upon
processes affecting plant growth (after Chapin 1983).
Even though arctic ecosystems are characterized by low nitrogen availability,
exceptionally rich areas exist under coastal cliffs, where nesting birds transport
nutrients from the sea to the land (Odasz 1994). In our study we chose to compare
vegetation in these nutrient rich areas with vegetation in nutrient poor areas in similar
temperature regimes. We also compared nutrient poor areas with relatively high
temperatures with nutrient rich areas and low temperatures and areas with low
nutrient level but differing temperatures. The aim of our investigation was to show ­­
by using this “natural experiment” ­­ whether nutrients per se are limiting growth of
different arctic plant species, or whether the observed growth limitation is solely
attributed to low temperature. We predict that nutrient addition is able to compensate
for some of the growth constraints imposed by low temperature. Hence we assume
that part of the growth limitation in cold regions is in fact a low temperature induced
nutrient limitation.
74

Materials and methods
Study localities
This study was undertaken on Svalbard, an archipelago in the high arctic. 10 different
geographical regions (table 1) in the western and northern parts of the archipelago
were chosen as study areas. In this way a wide range of biotic and abiotic conditions
was covered, in particular all of the three bioclimatic zones present on Svalbard.
Localities were assigned to bioclimatic zones according to Elvebakk (1999). For the
region Florabukta this division was problematic as we observed Salix polaris growing
there. This region should, according to Jónsdóttir (2005), therefore have been placed
in bioclimatic zone B. However, Salix polaris also occurred in Lady Franklinfjorden,
a region that according to Elvebakk (2005) should be considered part of bioclimatic
zone A. Moreover the flora and landscape in Florabukta indicated a position closer to
bioclimatic zone A (pers. comm. D. F. Murray 2007) and we therefore assigned the
region to this zone. Where necessary, in terms of differences in environmental
conditions, each region was divided into different sites.
Table 1. UTM coordinates, bioclimatic zone, habitat, and mean 10cm soil temperatures for the
study localities.
Study locality (E/N) Bioclimatic
zone
(site code)
Florabukta (0571/8887)
Lady Franklinfjorden
(0588/8901)
Lågøya (NA)
Florabukta (0571/8888)
Mayerbukta (0478/8863)
Biscayarhuken (0448/8864)
Reinsdyrflya (0478/8863)
Fjortende juli breen (0433/8786)
Colesdalen (0502/8670)
Kapp Thordsen (0511/8709)
Ossian Sars (0445/8763)
Ossian Sars (0445/8763)
A(1)
A(2)
A(3)
A(4)
B(5)
B(6)
B(7)
B(8)
C(9)
C(10)
C(11)
C(12)
Habitat 10cm soil
temperature in
°C
Mesic L. nivalis
L. confusa polar desert
P. dahlianum polar desert
Birdcliff
Glacier foreland
Mesic Luzula confusa
P. alpina snow bed community
Birdcliff
Thermophilic heath
Heath and calcerous fen
C. tetragona heath
Birdcliff
75
4,1
3,3
not measured
4,1
6,9
Sampling and sample analysis
Biometric plant traits from the following vascular plant species were collected:
Bistorta vivipara, Cerastium arcticum, Cochlearia groenlandica, Dryas octopetala,
Juncus biglumis, Luzula confusa, Luzula nivalis, Micrantes nivalis, Oxyria digyna,
Papaver dahlianum, Poa arctica, Poa pratensis, Ranunculus pygmaeus, Ranunculus
sulphureus, Saxifraga cernua, Saxifraga rivularis, Silene acaulis and Taraxacum
brachyceras. These species were selected due to their wide distribution on Svalbard.
At each study site, maximum plant height and leaf size of plants belonging to our
focal species were measured with a ruler. Maximum plant height was defined as the
longest part of the plant measured from the base to the tip of either shoot, leaf or
4,7
4,4
10
5,5
7,1
9,3
8,2

flower. Care was taken to measure only flowering individuals to ensure that plants
were at the same phenological state. A phenotypic plasticity index (ppi) for each of
the studied species was defined as 1 – (the ratio of measured minimum
height/maximum height).
The study sites were characterized by soil moisture estimates and total vegetation
cover, harvesting aboveground biomass, and by measuring the pH and soil
temperature. The moisture content of the soil was estimated by squeezing soil
between the fingers and assigning the results to a four­point scale (sensu Raup 1969).
Total vegetation cover was estimated per 0.5 x 0.5 m 2 by eye. Aboveground biomass
was harvested in representative full cover 10 x 10 cm plots, dried and weighed. The
biomass data was then scaled up to g/m 2 . Soil temperature measurements were taken
at 10 cm depth and were used as indicators for mid­summer temperature. These
values should be used with caution because of their sensitivity to variation in weather
conditions. But given the soil depth and 24 h day length (i.e. small diurnal amplitude),
the temperatures are somewhat buffered from short­term fluctuations and will
approximate the mean of the recent (few days) past with an accuracy of better than ±1
K. Mineral soil samples from each site were collected in the field and pH was
measured electronically in a laboratory. In cases where more than one value for the
abiotic factors mentioned above was obtained, the average value of these was used.
Finally the sites were divided according to a two­point nutrient level scale (1=nutrient
poor, 2=nutrient rich). All bird cliff sites as well as one site in Colesdalen close to an
old stable were assigned to nutrient level 2.
Statistical analysis
For the statistical analysis the statistics­software R 2.6.1 (R Development Core Team)
was used. Linear regression and ANCOVA were used to test for significance between
height/leaf size and temperature, nutrient level, moisture, pH, cover and bioclimatic
zone. The data was log­transformed to normalize distribution before running the
statistical tests.
Results
Biotic and abiotic factors
When all species were pooled and tested statistically, ANCOVA results showed a
positive correlation between plant height or leaf size and soil temperature as well as
nutrient level (Table 2, Fig 2). In addition these two analysis showed a significant
interaction between temperature and nutrient level (temperature x nutrients) in their
effect on plant height and leaf size. Plant height is also correlated with bioclimatic
zone, whereas leaf size is not. According to the results, moisture also has a significant
effect on plant height, but not on leaf size, but we cannot separate the direct moisture
effect from the indirect effect on soil development and nutrient availability.
Furthermore, a significant relationship between pH and plant height but not leaf size
was found. Vegetation cover has a significant effect on both plant height and leaf size.
The denser the plant community, the higher were the plants and the longer were the
leaves.
76

Figure 2. Overall plant height in areas with low nutrient availability (Low N) as well as areas
with high nutrient availability (High N) related to soil temperature in 10 cm depth.
Data points for means per species and site.
Table 2. Results of ANCOVA testing for significant correlations between several abiotic
parameters and overall plant height and leaf size among the studied sites. Bold font =
only single factors in the model, asterisk = model including interaction, NS = not
significant.
Plant height Leaf size
Parameter F df p F df p
Temperature
Nutrient level
temp* nut
Bioclimatic zone
Soil moisture
Soil pH
Vegetation cover
197.8 1,761

Biomass
Biomass per unit land area of full cover plots declined slightly from bioclimatic zone
C to A, and bird cliffs showed much higher biomass than nearby reference sites. The
relative fertilizer effect seamed to decrease from zone C to A, but the low number of
replicates did not allow a statistical test of this (Fig. 3).
Figure 3. Mean biomass in g d.m. m ­2 grouped into bioclimatic zones (A, B and C, i.e. from
cold to less cold). The means for each region lach data for region 1, 3 and 6. Data
from bird cliffs in black.
Biomass was nearly always >100 gm ­2 , and in two cases from bird cliffs 500­600 gm ­
2 , with the length of the growing season ranging from ca. 40 days (bioclimatic zone A)
to 80 days (bioclimatic zone C).
Phenotypic plasticity
The phenotypic plasticity index (ppi) for each of the studied species is shown in Fig.
4. The plasticity index calculated on the basis of measured minimum and maximum
leaf size shows a similar pattern like that in plant size and is therefore not shown here.
High values for the plasticity index indicate high variation in maximum plant height.
The mean plasticity value obtained from the measurements in this study is 0.83.
Species with relatively high variation in plant height are: Bistorta vivipara (ppi =
0.91), Cerastium arcticum p(pi = 0.96) Cochlearia groenlandica (ppi = 0.99), Oxyria
digyna (ppi = 0.96), Ranunculus sulphureus (ppi = 0.97), Ranunculus pygmaeus (ppi =
0.90) and Saxifraga cernua (ppi = 0.88) (Fig.4). Except for the Ranunculaceae, all of
these species are found among the tallest plants (Table 4).
78

1,00
0,95
0,90
0,85
0,80
0,75
0,70
0,65
0,60
0,55
0,50
B. vivipara
C. arcticum
C. groenlandica
D. octopetala
J. biglumis
L. confusa
L. nivalis
O. digyna
P. dahlianum
P. arctica
P. pratensis
79
R. pygmaeus
R. sulphureus
S. cernua
M. nivalis
S. rivalis
S. acaulis
S. uralensis
T. brachyceras
mean plasticity
plasticity index
Figure 4. Phenotypic plasticity index (1­min height/max height) for the studied plant species.
The line indicates mean ppi. The two species with the lowest index (

Luzula nivalis
Juncus biglumis
Dryas octopetala
Ranunculus pygmaeus
Silene acaulis
Saxifraga rivularis
114
111
70
70
55
25
80
0.38 1.10
1.33 1.37
0.63 1.02
1.18 2.35
Figure 5. Plant height of Bistorta vivipara related to soil temperature in 10 cm depth in areas
with low nutrient levels (low nutrient) compared to areas with high nutrient levels
(high nutrient).
Discussion
As expected there was a positive correlation between the three variables plant height,
leaf size, and biomass with temperature, as well as nutrient availability. Plant height,
leaf size, and biomass also increased from bioclimatic zone A to C. Our data shows
also a significant interaction between temperature and nutrient availability in their
effect on plant growth. Moreover, the results indicate that in accordance with our
hypothesis, increased nutrient levels can compensate for the growth limiting effect of
temperature in bioclimatic zone B and C, but not so in zone A. However, the impact
of differential temperature and nutrient level on the growth response of the studied
plants was not uniform among the different species.
Opportunists are supposed to respond faster and to greater extent to changes in
environmental parameters. Odasz (1994) used measures of nitrate reductase activity in
different plant species as an indicator for their ability to utilize nitrogen (Table 4). In
her ranking Oxyria digyna, Cochlearia groenlandica and Cerastium arcticum had a
high ability to utilize nitrogen. This should facilitate fast growth response when
nitrogen availability increase, a reaction that is especially pronounced in our data for
Oxyria digyna and Cochlearia groenlandica. Our results indicate that both Oxyria

digyna, Cochlearia groenlandica, but also Cerastium arcticum have high phentotypic
plasticity in response to nutrient availability, so this study supports the data and
suggestions by Odasz (1994).
Where our data allowed us to compare plants from nutrient rich localities in
bioclimatic zone B and C with individuals from nutrient poor sites in the same regions
we detected increased height in the nutrient rich sites. This finding can be interpreted
as a compensating effect of high nutrient levels for moderately low temperatures.
However, our data from nutrient rich and nutrient poor sites in bioclimatic zone A
does not show this pattern. There were strong signs of grazing by reindeer and geese
in the area (dune), but since we sampled the tallest undamaged individuals, this should
not have affected our height data. The lack of a clear response of plant height to
increased nutrient availability in bioclimatic zone A could also relate to a priority for
reproduction rather than vegetative growth. Arft et al. (1999) supposed that high arctic
species would respond to more favorable conditions in terms of temperature and
nutrients with an increase in reproductive effort rather than more growth (Baldeweg
and Bengtsson, AB­326 2007).
The fact that individuals of 11 out of 19 species, when growing in nutrient poor sites,
seemed to reach their maximum height in areas that did not have the highest measured
temperatures during our sampling is surprising. One explanation could be that denser
vegetation reduces soil heat flux. Another possible explanation could be unexpected
nutrient rich conditions at the moderate temperature site in Mayerbukta, indicated not
only by the observed higher plant growth but also by the high Shannon index (i.e.
high species diversity) observed at this site (Antonsson, Jørgensen and Tange, AB­
326 2007).
A problem in the analysis and interpretation of our data is the unbalanced replication
of the different species and the fact that many species did not occur in all
sites/regions. In particular, we lack sufficient data from N­rich sites (near bird cliffs)
for species that had been sampled elsewhere.
Conclusion
This study lends support to our hypothesis that high nutrient availability can
compensate for the growth limiting effect of low temperatures in the high Arctic as
long as temperatures are not too low. The nutrient addition effect seems to decline at
the coldest sites, which suggests that an ultimate (extreme) thermal limitation cannot
be mitigated by better nutrient supply.
Acknowledgements
Sincere thanks to Ane and Emma for their invaluable help with the statistics. Thanks
also to Inger Greve Alsos and Christian Körner for help in designing this project and
fruitful comments on the manuscript, as well as to our course mates for help with data
collection and a very nice time. Last but not least we would like to thank the crew on
“MS Stockholm”.
81

References
Arft, A. M., Walker, M. D., Gurevitch, J., Alatalo, J. M., Bret­Harte, M. S., Dale, M,
Diemer, M., Gugerli, F., Henry, G. H. R., Jones, M. H., Hollister, R. D., Jonsdottir,
I. S., Laine, K., Levesque, E., Marion, G. M., Molau, U., Molgaard, P., Nordenhall,
U., Raszhivin, V., Robinson, C. H., Starr, G., Stenstrom, A., Stenstrom, M., Totland,
O., Turner, P. L., Walker, L. J., Webber, P. J., Welker, J. M., Wookey, P. A. 1999:
Responses of tundra plants to experimental warming: Meta­analysis of the
international tundra experiment. Ecological Monographs 69, 491­511
Billings, W. D. 1987: Constraints to plant growth, reproduction, and establishment in
arctic environments. Arctic and Alpine Research 19, 357­3
Chapin III, F. S. 1983: Direct and indirect effects of temperature on arctic plants.
Polar Biology 2, 47­52
Chapin III, F. S. & Shaver, G. R. 1985: Growth response of tundra plant species to
environmental manipulations in the field. Ecology 66, 564­576
Elvebakk, A. 1999: Bioclimatic delimitation and subdivision of the arctic. In Nordal I
& Razzhivin VY. The species concept in the high north – A panarctic flora initiative.
Det Norske Videnskaps­kademi, Oslo
Elvebakk, A. 2005: A vegetation map of Svalbard on the scale 1:3.5 mill.
Phytocoenologia 35, 951­967
Elven, R., Murray, D. F., Razzhivin, V. & Yurtsev, B. A. Checklist of the Panarctic
Flora (PAF). Vascular Plants (Univ. of Oslo, Oslo, Norway, 2007)
Henry, G. H. R., Freedman, B. & Svoboda, J. 1986: Effects of fertilization on three
tundra plant­communities of a polar desert oasis. Canadian Journal of Botany 64,
2502­2507
Jonasson, S., Havstrøm, M., Jensen, M. & Callaghan, T. V. 1993: In­situ
mineralization of nitrogen and phosphorus of arctic soils after perturbations
simulating climate­change. Oecologia 95, 179­186
Jónsdóttir, I. S. 2005: Terrestrial ecosystems on Svalbard. Heterogeneity, Complexity,
and Fragility from an Arctic Island Perspective. Biology and Environment:
Proceedings of the Royal Irish Academy 105B, 155­165
Körner, C. 1989: The nutritional status of plants from high altitudes. A worldwide
comparison. Oecologia 81, 379­391
Körner, C. 2003: Alpine Plant Life. Functional Plant Ecology of High Mountain
Ecosystems. Springer­Verlag, Berlin Heidelberg
Odasz, A. M. 1994: Nitrate reductase­activity in vegetation below an arctic bird cliff,
Svalbard, Norway. Journal of vegetation science 5, 913­920
R Development Core Team 2007: R: A Language and Environment for Statistical
Computing. Vienna, Austria. http://www.R­project.org
Raup, H. M. 1969: The relation of the vascular flora to some factors of site in the
Mester Vig district, northeast Greenland. Meddelelse om Grønland, 176: 1­80
82

Reproductive allocation in the high Arctic
Heike Baldeweg 1 and Emma Bengtsson 2 , project leader Christian Körner 3
1 Institute of Ecology, Friedrich­Schiller­University Jena, Dornburgerstraße 159, D­07743, Germany
helia.antha@gmail.de, 2 Institute of Biology, NTNU, NO­7491 Trondheim, Norway
bengtson@stud.ntnu.no, 3 Institute of Botany, University of Basel, Schönbeinstrasse 6, 4056 Basel,
Switzerland, ch.koerner@unibas.ch
Abstract
From a given pool of resources, plants need to invest in vegetative or reproductive
structures, with obvious trade­offs between the two. In cold climates, plants are
generally smaller than in warm climates, and as life conditions become more adverse
miniaturization of plants progresses. Does this affect vegetative and reproductive
structures similarly (isometric allocation), or do plants give priority to one of the two
(asymmetric allocation), which would hint at selective pressure (evolutionary
adaptation)? In the high Arctic this has never been explored. Since sexual
reproduction through out­crossing increases the genetic variation and enables a
species to adapt to changing environments, we hypothesize that arctic plants prioritize
on reproduction, as has been shown for increasing altitude in the Alps. We tested this
hypothesis in Svalbard in summer 2007 along a climatic gradient between 78° N and
80° N. By using ratios between inflorescence size vs. plant height or leaf length we
found that forbs on average invest proportionally more in reproductive than vegetative
growth as it gets colder. In the graminoids studied along the same gradient, size
changes were isometric. We explain this by the different mode of sexual reproduction.
Insect pollination dependent species have to attract pollinators in an increasingly
barren landscape with overproportionally showy flowers. The trend we observed in
insect pollinated species matches the trend seen in the Alps. We may thus conclude
that arctic and alpine forbs prioritize dry matter allocation in favor of sexual
reproduction. Similar to the Alps, this does not hold for the wind pollinated
graminoids.
Introduction
The way in which plants reproduce determines the evolutionary processes and thus
the plants’ ability to respond to their environment. In alpine environments it has been
found that the allocation of resources changes towards the reproductive structures as
altitudes increases (Fabbro & Körner 2003). The Arctic is often compared with high
alpine environment assuming that there can be seen similar patterns in reproductive
allocation, but this has never been investigated.
With increasing latitude, as well as with increasing altitude, plants experience a
similar change in environmental conditions such as decreasing mean temperatures, a
shortening of growing season and a reduction in nutrient availability (Körner 2003). It
is widely known that plants living in cold climates adjust their morphology to better
cope with the predominant local conditions (Körner et al. 1989). Plants transplanted
from high to low altitudes and from high to low latitude tend to grow bigger and
83

partly lose their high altitude/latitude characteristics, indicating a high phenotypic
plasticity (Prock & Körner 1996).
Plant growth is almost always limited by some resources and once acquired, these are
invested into below and above ground structures. The above ground structures are
normally divided into the vegetative, photosynthesizing parts and the reproductive
parts or shoots (Seger & Eckhart 1996). Vegetative parts are the basis for further
resource acquisition and for growth and storage. The fraction of resources allocated to
reproductive parts does not produce such returns, but exerts costs. Hence there is a
trade­off between investment in structures which enhance vegetative growth versus
those which enhance reproduction.
There are three main strategies for plants to cope with this trade­off in cold climates
(Jonsdóttir 2005, Körner 2003):
1. Enhanced sexual reproduction: Resource investment to gametes, seeds and
seed dispersal
2. Clonal reproduction: Growth of vegetative structures that detach and grow into
new individuals with exactly the same genome as the mother plant
3. Space­holder strategy: maintain a position as long as possible (long life)
2 and 3 also need some sexual reproduction in the long run, but it is not essential on a
year­to­year basis.
A short growing season is disadvantageous for many flowering plants. It gives the
plant a very limited time span to develop blossoms, get pollinated and set seeds. In
addition, the short season limits seedling establishment. A shift to an alternative
strategy such as clonal propagation might be favourable. Billings (1974) proposed that
clonal propagation is an important advantageous factor in arctic environments. Clonal
reproduction (e.g. bulbils, runners and stolons) competes with sexual reproduction for
resources. Interconnected clones are more robust against disturbance and harshening
climates, because connected ramets can share resources ( Jonsdottir & Callaghan
1990, Carlsson & Callaghan 1990).
It has been suggested that sexual reproduction is the predominant, sometimes
the only reproductive mode, in as many as 60% of the 163 known Svalbard flowering
plants (Brochmann & Steen 1999). Sexual reproduction secures high rates of
evolutionary adaptation and, thus, better survival in stress­dominated environments,
i.e. greater fitness (Raffl et al. 2006). However, this needs outbreeding reproductive
systems which require cross­pollination, hence insect visitation. As a result of pollen
(pollinator) limitation, many high arctic species are shown to be highly selfcompatible
(Brochmann & Steen 1999). Selfing (fertilization of egg with pollen from the same
individual) has many disadvantages. The main disadvantage is inbreeding that reduces
heterozygosity. The loss of heterozygosity increases the expression of deleterious
alleles leading to inbreeding depression (Heshel et al. 2005). As negative this might
seem, inbreeding still generates some genetic variation because gamets are never
genetically identical (as result of meisosis), hence there is still a mixing of alleles in
each gamete relative to the other. This yields a small genetic diversification of the
offspring, which still might select for a better ability to cope with specific
environmental conditions (Martin & Hospital 2005).
84

Plants producing superior offspring are more widespread and have a higher over all
success. As long as there are events of insect pollination, the cost of investing in
reproductive organs is outweighed by the benefit of out crossing. Therefore, we
expect plants living in the high Arctic to invest more in their sexually reproductive
traits as Fabbro & Körner (2003) found in their comparison of lowland and alpine
plants in the Swiss Alps. They found that blossom­display area did not changed
between lowland and high alpine areas. At the same time they found that shoot size is
reduced nearly 10­fold. One of our objectives with this study was to see whether the
same can be found in the high Arctic, using a bioclimatical gradient in the archipelago
of Svalbard.
We hypothesize that high arctic plants, similar to alpine plants, invest relatively more
in reproductive structures than temperate low­land plants. In addition we expect that
within the arctic zone there is a trend towards greater allocation to inflorescences as
the climate gets harsher. In summary, we expect that
1. plants will become smaller in harsher climates
2. blossom diameter will not differ between bioclimatic zones
These predictions should hold for forbs, i.e. non­wind pollinated species. Whether this
holds for wind pollinated species depends on whether inflorescence size has similar
effects on capturing wind­driven pollen flow, which we will explore as well.
Materials and methods
A total of 37 species have been studied in ten different localities (see table 1) on the
northwest coast of the Svalbard archipelago. Svalbard spans over several degrees of
latitude and therefore has a wide range of temperatures. In addition the golf stream
affects climatic conditions at the west coast . In this investigation we followed the
definition of bioclimatic zones (BZ) by Elvebakk et al. (1999). Accordingly, places
with a mean July temperature between 0° and 2.5°C belong to BZ A, 2.5 to 4°C to B
and 4 to 6°C to BZ C.
In total 37 species were measured, the majority of them in at least two BZ. Seventeen
species have been measured in BZ A (localities 6, 7 and 8), 19 species in BZ B
(localities 2, 3, 4 and 5) and 23 species in BZ C (localities 1, 9 and 11).But because
the changes between the BZ’s that were interesting only use data for nine species (five
forbes and four graminoids) which were found in all the three BZ. Each species of
these were measured 15 to 95 times, in total 353 measurements.
In the Svalbard flora, most forbs can be divided into to major growth forms: cushion
plants and rosette plants. In the rosette plants the vegetative and reproductive organs
are not as easy to distinguish as is the case of cushion plants. The leaf rosette makes
up the vegetative part whereas the blossom is the reproductive part, but the stem often
shows photosynthetic activity, sometimes has leaves, but it is also the instrument of
exposing the blossom to the pollinating vector. The reproductive parts can be clonal
or sexual and many species, e.g. Saxifraga cernua, have both. For our study we
decided to measure only species in which individuals were easy to distinguish and
85

preferably grow in rosettes, which excludes cuhions. Cerastium arcticum had to be
included to enhance our dataset, even though it is mat­forming.
Table 1. Locality description.
Locality Latitude Longitude Characteristics (bedrock; vegetation type) Bioclimatic
zone
9 Ossian Sarsfjellet 78°93’ 12°47’ acidic to alkaline circumneutral;
Cassiope tetragona heath
C
10 Colesdalen 78°11’ 15°13’ weakly acidic; Dryas­Cassiope heath C
1 Kapp Thordsen 78°46’ 15°51’ weakly acidic; calcareous fens C
2 Mayerbukta 79°26’ 12°16’ acidic moraine B
3 Fjortende Julibreen 79°13’ 11°86’ acidic; mesic Luzula confusa B
4 Biskayabukta 79°84’ 12°40’ acidic; mesic Luzula confusa B
5 Reinsdyrsflya 79°83’ 13°90’ slightly acidic; Poa alpina snow bed comm. B
6 Florabukta 80°04’ 18°70’ alkaline circumneutral; mesic Luzula nivalis A/B
7 Nordre
Franklinbreen
80°15’ 19°66’ acidic; Luzula confusa polar desert A
8 Lågøya 80°21’ 18°17’ weakly acidic; Papaver polar desert A
Only plants that were mature but not senescent were measured. Measurements with a
ruler (to 1mm accuracy) included aboveground plant height, longest leaf length
(without petiole) or width, respectively (if the width was larger than the length) and
largest blossom diameter or length of the whole inflorescence in graminoids,
respectively. At least three samples of each forb species were harvested. The largest
single individuals from each species were chosen in each locality. The three plants
harvested for biomass determination were separated into stem fraction, leaf fraction
(including petioles) and blossom fraction, dried and weighed (only forbs, not
graminoids).
The blossom area (A) of forbs was calculated from diameter by assuming a circular
shape. A Person product­moment correlation was used to test the correlation between
biomass and biometric data. Differences in the ratios of height­leaf, blossom­leaf and
blossom height versus zones were tested with (one way) ANOVA. The same analysis
was taken to test differences in the biomass fractions vs. zones. In these statistical
tests we chose the ten biggest individual plants when we had more than ten
individuals. For the comparison between the bioclimatic zone­specific general means
across species were used. For the comparisons at the species level we used mean
86

values per species. A t­test was used to compare the graminoids of the Alps with the
graminoids of Svalbard. For all tests the program S­Plus 7.0 was used.
Results
Forbs
The correlation between biomass and biometrics measures was high for all traits
(blossom area p=0.013, R 2 =0.52; height: p=0.023, R 2 =0.71; leaf length p=0.016,
R 2 =0.68). Thus, for further analyses the biometric data for which we have larger
sample size were used only (rather than the biomass data).
For forbs, blossom size to leaf length ratio is significantly larger in harsher climates.
Also the blossom area to plant height ratio increases with harsher climate, but the
plant height to leaf ratio shows the opposite result and is decreasing (Table 2). In
other words, plants invest relatively more in blossoms when the climate gets more
adverse, and plant size is affected more than leaf size, which commonly means that
blossoms are displayed closer to the ground (Fig. 1). For individual species, the
blossom to plant height ratio was higher in BZ A than in C for most species (Table 3).
Figure 1. Reproductive allocation of plants in Svalbard. Inflorescence size corresponds to
blossom diameter in forbs and to spike length in graminoids. The temperatures are
according to Walker et al 2005 for the total arctic. The according temperatures for
Svalbard (oceanic climate are somewhat lower according to Elvebakk 1999, see
above.
87

Forbs
Graminoid
Graminoids
In graminoids only the plant height/leaf length ratio shows a decrease in harsher
climates (Table 2 and 4), which means that total plant height is more reduced than leaf
length. The inflorescence length/plant height ratio is unaffected (Fig. 1).
Table 2. Biometric ratios of forbs and graminoid species from the three bioclimatic zones on
Svalbard. * = p

Table 4. Species level comparison of ratios in graminoids in different bioclimatic zones on
Svalbard. No significant differences were found.
Zone A
mean±SE (n)
Zone B
mean±SE (n)
89
Zone C
mean±SE (n)
F­value
inflorescence length/leaf
length ratio
Juncus biglumis 0.23±0.074 (10) 0.15±0.05 (2) 0.236±0.036 (10) 0.41
Luzula confusa 0.12±0.013 (46) 0.24±0.02 (34) 0.232±0.021 (15) 21.47
Luzula nivalis 0.27±0.075 (6) 0.29±0.04 (4) 0.202±0.023 (10) 1.57
Poa arctica 0.76±0.11 (21) 0.93±0.13 (11) 1.256±0.098 (14) 6.27
inflorescence length/plant
height ratio
Juncus biglumis 0.11±0.01 (10) 0.09±0.02 (2) 0.117±0.017 (10) 0.15
Luzula confusa 0.08±0.01 (46) 0.14±0.01 (34) 0.106±0.016 (15) 12.52
Luzula nivalis 0.15±0.04 (6) 0.30±0.05 (4) 0.104±0.006 (10) 14.09
Poa arctica 0.35±0.02 (21) 0.40±0.02 (11) 0.386±0.029 (14) 1.20
plant height/leaf length
ratio
Juncus biglumis 1.83±0.35 (10) 1.61±0.18 (2) 2.097±0.224 (10) 0.65
Luzula confusa 1.59±0.09 (46) 1.66±0.08 (34) 2.697±0.367 (15) 6.91
Luzula nivalis 1.80±0.12 (6) 0.98±0.048 (4) 1.934±0.178 (10) 12.89
Poa arctica 2.20±0.315 (21) 2.26±0.28 (11) 3.429±0.280 (14) 5.73
Discussion
In our investigation we found that forbs allocated resources in favour of reproductive
structures as it gets colder. As predicted we found a significant reduction in overall
plant size and leaf size as climate gets harsher. However, blossom size of forbs
declined much less whereas graminoid inflorescences declined more or less
proportional to plant size (isometrically). Hence our hypothesis (2) was not falsified
for forbs, but it was falsified for graminoids.
On species level blossom area did not change in a homogenous way. Our results were
either non­significant, humped or showed a decrease towards BZ A. For P. dahlianum
and R. pygmaeus the hump might depend on higher competition in BZ C or optimal
habitat conditions in BZ B. We also have to consider the very small sample sizes for
these species in BZ C. In S. cernua we see the same decrease towards BZ A as in the
overall species comparison. In BZ C we measured only two plants of S. cernua and
they might not be representative, since they also reproduces asexually in form of
bulbils. Therefore it is difficult to draw valid conclusions about resource allocation in
this species.
There are few insect species and no bumblebees or bees in Svalbard (Coulson et al.
2003). Hence, the species assemblage in Svalbard lacks the species that are the most
important pollinators in temperate regions. This may enhance the evolutionary
pressure to produce showy flowers.
Diptera function as the main pollinators (Coulson et al. 2003). As most of them are
small, they have to stay close to the ground to avoid getting blown away by strong
winds that are typical for the archipelago. It seems that there is no reason for high

latitude plants on Svalbard to keep a long stem for attracting pollinators. Rather it is
the other way around. If the blossom gets closer to the ground they get closer to their
pollinators. Even if there were pollinators that might respond to taller flowering
stems, frost damage and mechanical damage by strong winds could impose higher
costs than the plant would benefit from exposing the blossom to potential pollinator.
Graminoid plant species are mostly pollinated by wind. If a wind pollinated plant is
not tall enough, the pollen falls on the ground, hence reducing dispersal. Having the
soike sheltered below the surrounding vegetation, pollen would not be captured by
wind. Our data illustrate the aerodynamic demands of graminoids and hint at different
evolutionary forces in forbs and graminoids.
A comparison with the data published by Fabbro & Körner (2003) indicates a
significant difference only in the blossom mass with a smaller mean blossom size in
the Svalbard archipelago, but in both cases blossoms are relatively larger in cold areas
(higher altitude or higher latitude, Figure 2). However, the effect is more pronounced
in the Alps (relatively bigger blossoms). Statistically tests of this comparison were not
possible, since we did not have sufficient data from the Alps. The rough estimate is
based on Figure 3 in the publication of Fabbro & Körner (2003).
The graminoid data collected in Svalbard were compared with the graminoid data
from the Alps (pers. communication by B. Krummen, Basel). T­tests showed no
difference in height and no difference in inflorescence length between the alpine belt
in the Alps and the high Arctic. The ratios show some difference though, with
inflorescence length­leaf ratio being statistically different (t = ­2.086, df = 45, p­value
= 0.0427). The inflorescence length plant height ratio does not differ and neither does
the plant height­leaf length ratio. Since leaf size measures were done differently in the
Alps, this is not surprising. Still the general conclusion for graminoids is that there is
no divergence between the Alps and the high Arctic. In both cases, graminoids and
their inflorescences decline in size isometrically.
The reason why we see such contrasting patterns between the two growth forms in
this comparison might be due to the different strategies of pollination, by wind or
animals. Wind dispersal probably does not differ that much between the high Arctic
and high altitudes in the Alps. On the other hand the insect pollinated species
experience a completely different set of pollinators that changes the selective pressure
on blossom morphology towards increasing display area in the Alps.
90

Blossom
Stem
Leaf
17,1%
20,2%
24,6%
33,8%
91
49,1%
55,2%
0,000 0,100 0,200 0,300 0,400 0,500 0,600 0,700
High latitude species High alpine species
Figure 2. Plant dry biomas fractions from the Alps and Svalbard (from Fabbro and Körner
2003). Biomass data from Svalbard not shown in the previous text, because they
show the same patterns as the size data.
Conclusion
As the general plant size decreases with harshening climate, the reproductive
investment is enlarged in forbs. Graminoids on the other hand, change isometrically
along the same gradient. This is probably due to different pollination systems. The
comparison between the Alps and Svalbard revealed that the blossom size relative to
vegetative traits in forbs is bigger in the Alps than in Svalbard, but the trend goes in
the same direction. Graminoids show the same patterns in Svalbard as in the Alps.
Pollinator­driven pollination thus selects inflorescence size differently from wind­
driven pollination.
Further studies in Svalbard on reproductive allocation should be more related to the
pollinating species, the possibility of selfing as well as the ability of clonal
reproduction.
For comparison between the alpine and arctic regions same genus should be
considered to reduce effects of basic morphological differences.
References
Billings, W. D. 1974: Adaptations and origins of alpine plants. Arctic and Alpine
Research 6:129­142
Brochmann, C. & Steen, S. 1999: Sex and Genes in he Flora of Svalbard­
Implications for Conservation Biology and Climate Change. Det Norske
Vitenskaps­Akademi. I. Matematisk Naturvitenskapelig Klasse, Skrifter, Ny Serie
38: 33­72
Carlsson, B. A. & Callaghan, T. V. 1990: Effects of flowering on the shoot dynamics
of Carex bigelowii along an altitudinal gradient in Swedish Lapland. Journal of
Ecology 78:152­165
Coulson, S.J., Hodkinson, I. D. & Webb N. R. 2003: Aerial dispersal over a high
arctic glacier foreland: Midtre Lovénbreen, Svalbard. Polar biology 26: 530­537

Elvebakk A. 1999: Bioclimatic delimitation and subdivision of the arctic. Det Norske
Vitenskaps­Akademi. I. Matematisk Naturvitenskapelig Klasse, Skrifter, Ny Serie
38: 81­112
Elvebakk, A. 2005: A vegetation map of Svalbard on the scale 1 : 3.5 mill.
Phytocoenologia 35: 951­967
Fabbro, T. & Körner, C. 2004: Altitudinal differences in flower traits and
reproductive allocation. Flora 199: 70­81
Heschel, M. S., Hausmann, N. & Schmitt, J. 2005: Testing for stress­dependent
Inbreeding Depression in Impatiens capensis (Balsaminaceae). American Journal of
Botany 92: 1322­1329
Jonsdottir, I. S. & Callaghan, T. V. 1990: Intraclonal translocation of ammonium and
nitrate nitrogen in Carex bigelowii Torr. ex Schwein. using 15N and nitrate
reductase assays. New Phytologist 114: 419­428
Jónsdóttir, I. S. 2005: Terrestrial ecosystems on Svalbard: heterogeneity, complexity
and fragility from an arctic island perspective. Biology and Environmen:
proceedings of the Royal Irish Academy, 105B: 155­165
Körner, C., Neumayer, M., Pelaez Menendez­Riedl, S. & Smeets­Scheel, A. 1989:
Functional morphology of mountain plants. Flora 182: 353­383
Körner, C. 1999: Alpine plant life­ Functional Plant Ecology if High Mountain
Ecosystems­ Springer Berlin
Martin, O. C. & Hospital, F. 2005: Two and Tree­Locus Tests for Linkage Analysis
Using Recombinant Inbred Lines. Genetics Society of America 173: 451­459
Prock, S. & Körner, C. 1996: A cross­continental comparison of phenology, leaf
dynamics and dry matter allocation in arctic and temperate zone herbaceous plants
from contrasting altitudes. Ecological Bulletins 45: 93­103
Raffl, C., Marcante, S. & Erschbamer, B. 2007: The role of spontaneous selfing in the
pioneer species Saxifraga aizoides. Flora 202: 128­132
Seger, J. & Eckhart, V.M. 1996: Evolution of sexual systems and sex allocation in
plants when growth and reproduction overlap. Proceedings of the Royal Society of
London series b­Biological Sciences 263: 833
92

Freezing resistance in high arctic plant species of
Svalbard in mid­summer
Christian Körner 1,3 and Inger Greve Alsos 2,3
1 Institute of Botany, University of Basel, Schönbeinstrasse 6, 4056 Basel, Switzerland,
ch.koerner@unibas.ch, 2 UNIS ­ The University Centre in Svalbard, P.O. Box 156, 9171 Longyearbyen,
Norway, ingera@unis.no, 3 Results of the summer course Arctic Plant Ecology (AB 326) at UNIS with
13 students: Ane Christensen Tange, Christian Engebretsen Pettersen, Eike Müller, Elke Morgner,
Emma Kristina Bengtsson, Heike Baldeweg, Henrik Antonsson, Ingelinn Aarnes, Kathrin Bockmühl,
Marte Holten Jørgensen, Merete Wiken Dees, Simone Lang, Unni Vik
Abstract
It is well known that freezing temperatures put arctic and alpine plants at risk during
the growing period only. Here we explore responses of high arctic plant species to a
simulated early summer freezing event of ­7 °C. After gradual cooling, several hours
at target conditions and slow thawing over 21 hours, the 12 species tested exhibited a
broad damage spectrum from zero to 100%. We conclude that ­7 °C is a critical
temperature for many arctic taxa, and repeated freezing episodes like this would exert
major changes in species abundance.
Key words: Cold climate, frost, low temperature, Spitsbergen
Introduction
Freezing resistance is the first and most fundamental environmental filter plant
species have to pass to establish sustainably in periodically cold environments. The
taxa found in areas with periodic freezing temperatures have to be resistant, otherwise
they would not be there. However, freezing tolerance of plants has two aspects. One is
the loss of certain (above­ground) organs such as leaves and flowers, another one is
the complete extinction of an individual. The repeated occurrence of the first may,
however, pave the way to the second. Another facet of freezing tolerance is time,
which again has two dimensions, a short term and a long term. In the short term,
freezing tolerance goes through seasonal cycles of acclimation, with hardening and
dehardening associated with both photoperiod and experienced temperature (Sakai &
Larcher 1987, Heide 2005). Plants are commonly much more susceptible when they
are active, i.e. during the growing season, than during the dormant season (winter).
Tissues of arctic and alpine plants may tolerate almost any temperature (some even
liquid nitrogen) when they are dormant, but might be killed at temperatures between ­
2 and ­9 °C, when actively growing and flowering. Hence it is well established that
extreme summer events are the critical issue to look at (for a review see Körner 2003).
The climate of Svalbard is quite untypical for the given latitudinal position between
76.5° and 80.5° of northern latitude, because the Archipelago is hit by the golf stream,
causing its temperatures to lack extremes known for other high arctic and antarctic
latitudes, both in terms of winter minima and summer maxima (Engelskjøn et al.
2003, Przybylak 2007). Although the annual mean temperature is ca. ­5 °C at sea
level at the Longyearbyen airport station (a comparatively warm area for Svalbard; for
93

a climate description see Rønning 1996), the 30 year absolute minima there are ­10 °C
for June and 0 °C for July (the lowest temperature ever measured, ­46 °C in March;
annual absolute maximum +23 °C in July). At sea level and in favorable regions, such
as the inner Fjords, the growing season normally does not last longer than 80­100
days in the warmest places. It may be as short as 6 weeks in less sheltered coastal sites
and there is hardly any higher plant growth in the polar desert north of 80°. Even the
most favorable locations do not permit upright shrubs to grow. The about 165 higher
plant species known for Svalbard (Brochmann & Steen 1999, Rønning 1996, Elven &
Elvebakk 1996), all have their growing points (apical meristems) within a few
centimeters above ground or below the soil surface (the majority), only exposing
leaves and flowers to atmospheric conditions. These tissues may thus be experiencing
short term low temperature extremes, while subsurface structures would profit from
the moderating influence of immersion in soil.
Although the sun is shining 24 hours in summer, and even midnight sun is quite
intense, slope inclination and slope direction can cause a periodic screening of
vegetation from direct insolation. While direct solar radiation is commonly causing
plant temperature to rise above air temperature in the arctic (by 4­17K according to
Biebl 1968), the lack of direct insolation can cause ground surface temperatures to
drop below air temperature by radiative cooling under clear sky conditions (Körner
2003). Temperatures may be 2­4 K below the air temperature a meteorological station
might record under standard measurement conditions. In addition, altitude causes a
mean reduction of 0.7K per100 m in this area and flowering plants grow above 500 m
a.s.l. in some places. Hence, plants may experience temperatures as low as ­15 °C in
June or ­5 °C in July, at least once in a century, at the climatic conditions recorded so
far. Most recently, the climate in Svalbard has warmed significantly, and most
glaciers retreat (Przybylak 2007). However, a warmer climate may not exclude/reduce
the likelihood of such extreme events, but may accelerate spring development and
thus, actually enhance the risk of plants becoming exposed to such conditions in June
(the beginning of the growing season). It would thus be good to know how sensitive
plant species actually are to summer time freezing events and whether such events
could cause differential damage across species and thereby influence community
composition. We explored this possibility in a small screening experiment, using fully
developed, flowering plant individuals of 12 common species growing at 30 m above
sea level, in Endalen 4 km east of Longyearbyen.
Methods
At least 5 mature leaves and 5 non­senescent, fully open inflorescences (except for S.
nivalis) were collected for each of three treatments (see below) from randomly
selected individuals around 6 p.m. on July 10, 2007 (3x5=15 samples for 12 species).
Samples were collected into and transported to the lab in 100 ml plastic vials and were
exposed to the following three conditions by 8 p.m. the same day: a cool room at +5
°C (control), a freezer set at ­18 °C (guaranteed freezing damage) and a freezer set to ­
7 °C. We had selected the ­7 °C treatment because this temperature falls midway the
known range of summer­time freezing damage in alpine plant species (Körner 2003).
All species analyzed here (Table 1) are widespread in Svalbard and classified as
temperature indifferent species except Dryas octopetala, which is classified as weakly
thermophilous, and Petasites frigidus and Silene furcata, which both are classified as
disinctly thermophilous species (Elvebakk 1989). Dryas octopetala is common in the
northern arctic tundra zone but does not grow in the polar desert zone. Petasites
94

frigidus and S. furcata are limited to the climatically more favorable inner fjord zone
of Svalbard (www.svalbardflora.net).
The two groups of vials in the freezers were placed in insulating boxes, so that the
lowest temperature was approached slowly (over several hours), persisted over several
hours (until noon of July 11) and was slowly allowed to ramp back to +5 C in the cool
room. Temperatures were logged inside the boxes with one channel temperature
loggers (±0.2 K, Tidbit, Onset Corp., Bourne, USA). By 5 p.m. samples were
removed from the cool room and immersed in distilled water in petri­dishes on white
paper at room temperature (Fig. 1). The next morning (12 July) samples were
inspected by eye for discoloration, turgor and/or smell and the immersion water was
checked conductometrically for ion leaching (cell membrane collaps). Because the
amount of tissue differed greatly (because of different leaf and flower sizes) the
conductometric readings have only semi­quantitative value. The visual inspection was
repeated 24 hours later. In flowers we counted the number of damaged flowers, in
leaves we estimated the % leaf area damaged per leaf, and calculated a mean.
Figure 1 a­b. Assessment of tissue conditions after freezing treatment in petri­dishes filled with
water.
95

Results and Discussion
The minimum temperatures recorded in the vicinity of the samples in the freezers
were ­18.2 and ­7.2 °C. All control tissue (+5 °C) remained intact, freshly looking,
and served as a reference. All ­18° treated tissue was dead and served as a reference
for how damaged tissue looks like. The ­7.2 treated cohort exhibited a clearly species
specific differentiation in damage, with some species/organ types completely
damaged or undamaged or partially damaged (Tab. 1, Fig.2). Two species turned out
to be difficult to assess. Saxifraga nivalis leaves largely looked fine, but there was
significant ion leakage to the immersion water. Luzula confusa naturally has reddish
and rigid leaves and flowers and neither showed clear color changes nor turgor loss,
nor was ion leakage obvious, but some leaves and flowers looked slightly altered. All
other species/organs could be clearly judged. Damage in flowers was generally much
greater, but our sample size was too small to rank species in a trustworthy manner.
We thus, present two damage categories only for flowers (Tab. 2).
Table 1. Ranking of examined plant species for freezing damage in leaves.
_____________________________________________________________________
_________
Rank Plant species Damage (%) Conductivity (µS)
1 Papaver dahlianum 0 19
2 Dryas octopetala 2 17
3 Petasites frigidus 7 51
4 Saxifraga nivalis 10 54
5 Salix polaris 12 17
6 Cerastium arcticum 26 114 a
7 Luzula confusa 40 19
8 Saxifraga cespitosa 58 ­ b
9 Oxyria digyna 60 138
10 Bistorta vivipara 84 63 c
11 Saxifraga cernua 98 239
12 Silene furcata 100 383
a Central, young part of rosette alife, outer and bigger part (older leaves) dead.
b Some soil and dead leaf material attached to rosettes, i.e. odd reading.
c Note, the small leaf area is causing a small signal despite large damage.
In summary, these data show that a ­7 °C freezing event at that time of the year and/or
developmental stage of plants would have a dramatic impact on above ground
structures of these plants in mid­summer. Yet, species known for their occurrence in
the most extreme polar habitats, such as Papaver dahlianum showed not only no leaf
damage, but even a fraction of the otherwise delicate looking, big flowers had
survived. Dryas octopetala commonly ranked as weakly thermophilous (Elvebakk
1989), survived with almost 100% intact foliage. Even more surprising, Petasites
frigidus, ranked as distinctly thermophilous (Elvebakk 1989), showed over 90 %
survival. Thus the distribution of this species is probably not limited by extreme low
temperatures. Rather, other factors such as limited ability for sexual reproduction
(Brochmann & Steen 1999, Rønning 1996) may limit its distribution in Svalbard.
96

Fig. 2 a­f. Leaf and flower conditions 30 hours after thawing. From left to right: control (+5
°C), “mild” freezing (­7.2 °C), severe freezing (­18.2 °C). a Silene and Cerastium, b
Papaver and Saxifraga cespitosa, c Bistorta and S. cernua, d Luzula and Dryas, e S.
nivalis and Salix, f Oxyria and Petasites
97

Table 2. Ranking of examined plant species by freezing damage in flowers a .
(1) 10­30% survival, i.e. 70­90 % damage:
Oxyria digyna
Dryas octopetala
Luzula confusa
Saxifraga cespitosa
Papaver dahlianum
Saxifraga nivalis b
(2) 0% survival, i.e. 100 % damage:
Silene furcata
Cerastium arcticum
Bistorta vivipara
Saxifraga cernua
Salix polaris
a Petasites frigidus had no flowers.
b Uncertain, flowers looked almost ok.
Of the 5 individuals (leaf rosette with inflorescence) of Saxifraga cespitosa, 2
individuals showed negative geotropism over night, i.e. they uprighted their fully
intact flowering shoots from the flat position in the petri­dish (Fig.3). For half of the
tested species, such an event would eradicate the complete reproductive investment of
that season, but in these species most or all leaves would also be killed.
Silene furcata, the other distictly thermophilous species analysed here, emerged as the
most sensitive species. Hence, it may serve as an indicator species of favourable site
conditions, provided such selective freezing would occur. The plant genus does not
seem to be of predictive value, given that Saxifraga exhibits particularly sensitive and
robust representatives. Silene acaulis was found to be very robust against midsummer
freezing near Tromsø (surviving ­8.5 to ­9 °C; Junttila & Robberecht 1993). The latter
authors had confirmed earlier evidence that growth temperatures have a strong
influence on actual freezing resistence. It was thus considered that repeated warm
episodes in the arctic could weaken freezing tolerance (Marchand et al. 2006). We
sampled plants during a very warm period, with noon air temperatures of around +12
°C, which could have sensitized our test plants.
The low freezing resistance of Saxifraga cernua was unexpected given that this
species is common in the polar desert zone and is one of the species reaching highest
up in the mountains in Svalbard (>900 m a.s.l., Sunding 1960). Also, some other
hardy arctic species such as Bistorta vivipara, Oxyria digyna and Saxifraga cepitosa
showed over 50 % freezing damage. Common for all these species is their ability to
produce viable seeds or bulbils even at cold sites (Cooper et al. 2004). Recruitment
from seed bank may therefore replace any lost individuals due to freezing events and
thus secure long­time survival of these species even at the coldest sites in the Arctic.
98

Fig. 3 a­c. Close­up examples of flower conditions after the ­7°C treatment. Note the negative
geotropic position of 2 Saxifraga caespitosa shoots after 30 hours (3a), flowers (one
of each damaged at
­7 °C and one undamaged at ­7 °C) of Dryas octopetala (3b) and Papaver dahlianum
(3c).
It can be concluded that during the growing season, freezing temperatures of ­7 °C or
possible less, would damage a significant fraction of the arctic flora. Given that ­10
°C had been recorded for air temperature in June, this means that earlier spring
growth in the course of global warming at an otherwise unchanged likelihood of the
occurrence of such an extreme events would lead to massive losses of tissue and
above ground productivity. It would be interesting to know, whether low temperature
extremes during the growing season had in fact became rarer as the overall means in
temperatures rose in recent years. Our survey indicates the critical range of
temperature to be explored in a more detailed assessment and that biodiversity
(species identity) matters a lot. Overall, the data appear to match the summer freezing
tolerance known for alpine plants of the temperate zone (Körner 2003) and do not
indicate a greater frost hardiness for these high arctic plants as had been suggested
from experiments with Saxifraga oppositifolia under controlled growth conditions
(Robberecht & Junttila 1992).
Acknowledgements
UNIS organized and financed this arctic plant ecology course in July 2007 and
provided the needed infrastructure. The participating students contributed the practical
work and are listed on the title page.
99

References
Biebl, R. 1968: Über Wärmehaushalt und Temperaturresistenz arktischer Pflanzen in
Westgrönland. Flora 157:327­354
Brochmann, C. & Steen, S. 1999: Sex and genes in the flora of Svalbard ­
Implications for conservation biology and climate change. Det Norske Videnskaps­
Akademi. I. Matematisk­ Naturvitenskapelig Klasse, Skrifter, Ny Serie 38:33­72
Cooper, E.J., Alsos, I.G., Hagen, D., Smith, F.M., Coulson, S.J. & Hodkinson, I.D.
2004: Recruitment in the Arctic: diversity and importance of the seed bank. Journal
of Vegetation Science 15: 115­124
Elvebakk, A. 1989: Biogeographical zones of Svalbard and adjacent areas based on
botanical criteria. Institute of Biology and Geology. University of Tromsø, Tromsø
Elven, R. & Elvebakk, A. 1996: Part 1. Vascular plants. A catalogue of Svalbard
plants, fungi, algae, and cyanobacteria (ed. by A. Elvebakk & P. Prestrud), pp 9­55.
Norsk Polarinstitutt, Oslo
Engelskjøn, T., Lund, L. & Alsos, I.G. 2003: Twenty of the most thermophilous
vascular plant species in Svalbard and their conservation state. Polar Research 22:
317­339
Heide, O.M. 2005: Ecotypic variation among European arctic and alpine populations
of Oxyda digyna. Arct Antarct Alp Res 37:233­238
Junttila, O. & Robberecht, R. 1993: The influence of season and phenology on
freezing tolerance in Silene acaulis L., a subarctic and arctic cushion plant of
circumpolar distribution. Ann Bot 71:423­426
Körner, C. 2003: Alpine plant Life. Springer, Berlin
Marchand, F.L., Kockelbergh, F., van de Vijver, B., Beyens, L. & Nijs, I. 2006: Are
heat and cold resistance of arctic species affected by successive extreme
temperature events? New Phytol 170:291­300
Przybylak, R. 2007: Recent air­temperature changes in the Arctic. Annals of
Glaciology 46:316­324
Robberecht, R. & Junttila, O. 1992: The Freezing Response of an Arctic Cushion
Plant, Saxifraga caespitosa L.: Acclimation, Freezing Tolerance and Ice
Nucleation. Ann Bot 70:129­135
Rønning, O.I. 1996: The flora of Svalbard. Norwegian Polar Institute, Oslo
Sakai, A. & Larcher, W. 1987: Frost Survival of Plants. Responses and Adaptation to
Freezing Stress. Ecological Studies 62, Springer, Berlin
Sunding, P. 1960: Høydegrenser for høyere planter på Svalbard (Height limits for
vascular plants in Svalbard). Norsk Polarinstitutt Årbok 32­59
100